Skip to main content
NCBI home page
Toxicology Research logoLink to Toxicology Research
. 2018 Apr 18;7(3):358–370. doi: 10.1039/c8tx00016f

Evaluation of the relevance of DILI predictive hypotheses in early drug development: review of in vitro methodologies vs. BDDCS classification†

Rosa Chan a, Leslie Z Benet a,✉
PMCID: PMC5959290  NIHMSID: NIHMS964288  PMID: 29785262

graphic file with name c8tx00016f-ga.jpgDrug-induced liver injury (DILI) is a major safety concern; it occurs frequently; it is idiosyncratic; it cannot be adequately predicted; and a multitude of underlying mechanisms has been postulated.

Abstract

Drug-induced liver injury (DILI) is a major safety concern; it occurs frequently; it is idiosyncratic; it cannot be adequately predicted; and a multitude of underlying mechanisms has been postulated. A number of experimental approaches to predict human DILI have been proposed utilizing in vitro screening such as inhibition of mitochondrial function, hepatobiliary transporter inhibition, reactive metabolite formation with and without covalent binding, and cellular health, but they have achieved only minimal success. Several studies have shown total administered dose alone or in combination with drug lipophilicity to be correlated with a higher risk of DILI. However, it would be best to have a predictive DILI methodology early in drug development, long before the clinical dose is known. Here we discuss the extent to which Biopharmaceutics Drug Disposition Classification System (BDDCS) defining characteristics, independent of knowing actual drug pharmacokinetics/pharmacodynamics and dose, can be used to evaluate prior published predictive proposals. Our results show that BDDCS Class 2 drugs exhibit the highest DILI severity, and that all of the short-lived published methodologies evaluated here, except when daily dose is known, do not yield markedly better predictions than BDDCS. The assertion that extensively metabolized compounds are at higher risk of developing DILI is confirmed, but can be enhanced by differentiating BDDCS Class 2 from Class 1 drugs. Conclusion: Our published analyses suggest that comparison of proposed DILI prediction methodologies with BDDCS classification is a useful tool to evaluate the potential reliability of newly proposed algorithms, although BDDCS classification itself is not sufficiently predictive. Almost all of the predictive DILI metrics do no better than just avoiding BDDCS Class 2 drugs, although some early data with microliver platforms enabling long-enduring metabolic competency show promising results.

Introduction

Drug-induced liver injury (DILI) remains a major safety concern “due to its frequency of occurrence, idiosyncratic nature, poor prognosis, and multiple underlying mechanisms. Numerous experimental approaches have been published to improve human DILI prediction with modest success”, as noted by Shah et al.1 Idiosyncratic DILI (IDILI) is very complex. Most IDILI appears to be immune mediated, and reactive metabolites appear to be involved in most, but not all IDILI. Reactive metabolites are widely accepted as playing a pivotal role in the pathogenesis of idiosyncratic adverse drug reactions. While there are today well-established strategies for the risk assessment of stable metabolites within the pharmaceutical industry, there is still no consensus on reactive metabolite risk assessment strategies.2 This is due to the complexity of the mechanisms of these toxicities as well as the difficulty in identifying and quantifying short-lived reactive intermediates such as reactive metabolites. In addition, there are probably several mechanisms by which a drug or reactive metabolite can induce an immune response. However, one might anticipate that new long-lived hepatocytes (i.e., liver-on-a-chip) might be better able to evaluate these conditions.

Across the pharmaceutical industry, systems of screening drug candidates have emerged that include transcriptomic profiling of animals in addition to animal pathology, assessment of covalent binding and glutathione (GSH) adducts in microsomal test systems and in vivo, inhibition of bile salt export pump (BSEP) in vitro, impairment of function of isolated animal mitochondria, and cell stress responses and viability in human hepatoma and hepatocyte culture systems. Several common themes emerge in all these test systems especially involving oxidative stress, mitochondrial impairment, covalent binding, and endoplasmic reticulum stress. It has been proposed that these test systems have moderately strong predictive value for IDILI,3–5 which we evaluate here. Others have examined combinations of mechanistic assays to better predict hepatotoxicity potential,2,6 also evaluated here. Several studies have shown a correlation of total administered dose alone7 or in combination with drug lipophilicity8 with higher risk of DILI. However, it would be best to have a predictive DILI methodology early in drug development, long before the clinical dose is known.

Since liver injury has been reported with a large number of drugs, efforts have been undertaken to compile human hepatotoxicity data, including the National Institute of Health LiverTox Database9 and the FDA Liver Toxicity Knowledge Base (LTKB).10 These publicly available datasets have enabled development of new structure activity relationships for hepatotoxicity endpoints or triggered the development of knowledge-based and quantitative structure activity relationship (QSAR) models.1,11,12

We have reviewed the applicability of the Biopharmaceutics Drug Disposition Classification System (BDDCS) compared with presently proposed predictive procedures in evaluating DILI toxicity. Since its inception, the BDDCS has been useful in drug discovery for predicting routes of elimination, oral drug disposition, food effects on drug absorption, transporter effects on drug absorption, and potentially clinically significant drug interactions that may arise in the intestine, liver and brain.13 We have shown that the BDDCS can be useful in predicting the potential for antiepileptic drugs to cause cutaneous adverse reactions14 and in predicting DILI.15 Most recently, we have shown that in vitro measures of BSEP inhibition, alone and together with other efflux transporters, provides no better prediction of DILI than just avoiding BDDCS class 2 drugs.16 BDDCS's strong relationship between dose, metabolic susceptibility, solubility and idiosyncratic DILI highlights the potential benefits of BDDCS as a comparison matrix for DILI prediction.

The BDDCS was developed in 2005 after Wu and Benet recognized that highly permeable compounds, as outlined by the Biopharmaceutics Classification System (BCS), were extensively metabolized, while poorly permeable drugs were primarily eliminated unchanged in the urine or bile.17 BDDCS demonstrated that simple passive membrane permeability measures were highly selective in differentiating extensively vs. poorly metabolized drugs in humans. Drugs in the BDDCS are classified according to the membrane permeability rate and aqueous solubility. These characteristics have helped BDDCS define whether metabolic enzymes and/or transporters are clinically important. BDDCS features are demarcated by high and low values, classifying drugs into four categories. These classes are each associated with specific predictions regarding route of elimination and which interactions may be a clinical concern.13

Here we provide a review on the extent to which BDDCS defining characteristics, independent of knowing actual drug pharmacokinetics/pharmacodynamics and dose, can be used as a comparison baseline matrix of potential DILI adverse events with prior published predictive proposals.13,15,16 We review the clinical impact of BDDCS in evaluating the severity of DILI warnings in drug labels approved by the Food and Drug Administration (FDA),18 the withdrawal status due to adverse drug reactions (ADRs), the role of BSEP inhibition, maximum daily dosages prescribed, and in vitro toxicology assays applied to cover various mechanisms and toxicity endpoints associated with human DILI.15

Assessment of the BDDCS classification on FDA drug labels associated with DILI hepatic liability

In our previous work, we reported the BDDCS class relationship of hepatotoxicity between the different ADR categories by calculating the proportion of drugs in each FDA hepatic liability category, and each DILI severity category.15 As depicted in Fig. 1A, we observe that as the hepatic warning severity increases, the proportion of BDDCS Class 2 drugs increases and the proportions of both BDDCS Class 1 and 3 drugs decrease, all with highly significant trends. The “No mention” category is significantly different from all other categories, except for “Adverse Reactions.” BDDCS Class 2 drugs exhibited the highest proportions in the following drug label sections: “Warning and Precautions” (45.6%, 36/79), “Boxed Warning” (47.2%, 17/36), “Withdrawn” (62.5%, 25/40) and “Discontinued” (83.3%, 5/6). Obviously, the number of drugs designated as exhibiting severe DILI increases as the ADR severity increases. That is, 15.9% (7/44) in the “Adverse Reactions” category, 36.7% (29/79) in the “Warning and Precautions” and 81.6% (31/38) of the drugs in the “Black Box Warning” are assessed to exhibit severe DILI.15 In Fig. 1B and C, the two BDDCS determinants (extent of metabolism and solubility) are examined. The percentages of poorly metabolized (Fig. 1B) and of highly soluble (Fig. 1C) drugs show statistically significant decreases with hepatic liability, while low solubility drugs increase significantly (Fig. 1C) with hepatic liability. The percent of extensively metabolized drugs also increases with hepatic liability, but since almost 2/3 of “No mention” drugs are metabolized, it is apparent that extent of metabolism itself is not a discriminating parameter. Although greater extent of metabolism has been reported to significantly increase the potential of a compound to cause DILI,19 this property alone is not able to distinguish compounds that are “No mention” of hepatic liability from those compounds exhibiting hepatic liability (see Fig. 1B).

Fig. 1. (A) Distribution of FDA hepatic liability with BDDCS Class (Reprinted with permission from R. Chan and L. Z. Benet, Evaluation of DILI predictive hypotheses in early drug development, Chem. Res. Toxicol., 30, 1017–1029. Copyright 2017 American Chemical Society.) Drugs were assigned according to the most severe drug label section reporting a hepatic ADR or Withdrawn and Discontinued, and to the “No mention” class if no hepatic ADRs were reported. Bars show the percentage of all compounds in the same category that are associated with each FDA hepatic liability. BDDCS Class 2 drugs are shown to significantly increase the frequency of hepatic ADRs reported in the “Warnings and Precautions”, “Boxed Warning”, “Withdrawn” and “Discontinued” categories. There was a significant difference between BDDCS Classes when the proportionality trend test was calculated: BDDCS Class 1 trend p-value = 0.0003842; BDDCS Class 2 trend p-value = 2.014 × 10–10; BDDCS Class 3 trend p-value = 0.003928; BDDCS Class 4 trend NS, p-value = 0.2963. Differences in the BDDCS Class distributions were evaluated among the following groups: “No mention” vs. “Adverse Reactions”, NS, p-value = 0.2908; “No mention” vs. “Warning and Precautions”, p-value = 0.0001058; “No mention” vs. “Boxed Warning”, p-value = 0.0006222; “No mention” vs. “Withdrawn”, p-value = 6.439 × 10–7; “No mention” vs. “Discontinued”, p-value = 9.36 × 10–5. (B) Distribution of FDA hepatic liability with respect to extensively metabolized vs. poorly metabolized drugs. The extensive hepatic metabolism group consisted of 102 BDDCS Class 1 and 99 BDDCS Class 2 drugs; the poor hepatic metabolism group consisted of 55 BDDCS Class 3 and 8 BDDCS Class 4 drugs. There was a significant difference between extensively metabolized vs. poorly metabolized drugs when the proportionality test was calculated (p-value = 0.001536). (C). Distribution of FDA hepatic liability with respect to high solubility vs. low solubility drugs. The high solubility group consisted of 102 BDDCS Class 1 and 55 BDDCS Class 3 drugs; the low solubility group consisted of 99 BDDCS Class 2 and 8 BDDCS Class 4 drugs. There was a significant difference between high solubility vs. low solubility drugs when the proportionality test was calculated (p-value = 3.481 × 10–9).

Fig. 1

Open in a new tab

Our examination of the relationship between the BDDCS's determinant properties: solubility and extent of metabolism, led to some novel observations. Drugs belonging to BDDCS Class 1 and 3 exhibited a lower proportion of DILI severity. Drugs that are extensively metabolized and have low aqueous solubility, i.e., BDDCS Class 2 drugs, have the highest rates of DILI risk. BDDCS Class 2 drugs exhibited the highest proportions among the “Warning and Precautions”, “Black Box Warning”, “Withdrawn” and “Discontinued” categories. These are notably considered the most serious DILI risk categories (see Fig. 1A). These findings demonstrate the importance of intrinsic drug properties as a potential factor for the development of a DILI event.

Drugs belonging to BDDCS Class 3 and 4 exhibited much lower proportions in the FDA hepatic liability (see Fig. 1A). Moreover, BDDCS Class 3 and 4 drugs show a decreased risk of DILI related pathological events, such as liver aminotransferases increase and hyperbilirubinemia. However, we note the underrepresentation of BDDCS Class 4 drugs in the overall scheme of marketed approved drugs. Compounds with poor hepatic metabolism had been previously noted to be significantly less likely to cause hepatotoxicity.19 Although a lack of hepatic metabolism does not assure total lack of hepatotoxicity, it indeed appears that BDDCS Class 3 and 4 drugs lead to a lower DILI severity.

Barton and co-workers20 have previously discussed a new paradigm for navigating compound properties related to drug attrition. Optimizing the exposure of potent compounds at the desired site of action and in tissues associated with toxicity is fundamental to addressing attrition via efficacy and safety. Traditional oral drug space is well defined with respect to physicochemical properties and absorption, distribution, metabolism, excretion and toxicity (ADMET) risks but increased focus on ligand-lipophilicity efficiency, maximizing enthalpy contributions and new target classes challenge this paradigm. Barton et al.20 propose that BDDCS Class 3 compounds should be significantly more associated with drug attrition because they tend to be transporter substrates or inhibitors. Furthermore, they suggest that compounds that are substrates for transporters as being a toxicity liability. We completely disagree with this suggestion based on our analysis of DILI potential15 and antiepileptic drugs’ cutaneous adverse events.14 Our analysis suggests that BDDCS Class 3 compounds exhibit less toxicity potential.

Assessment of daily dosage on FDA drug labels and DILI severity

Numerous compound- and/or patient-specific risk factors can contribute to the susceptibility to DILI. IDILI has been shown to be dependent on both daily dose and extent of hepatic metabolism of a drug.7,19,21,22 Lammert and coworkers7,19 have attributed hepatic adverse events to compounds with significant hepatic metabolism and daily dose ≥50 mg. Formation of reactive metabolites, high covalent body burden,23,24 mitochondrial dysfunction (resulting in the depletion of cellular energy supply and the generation of damaging reactive oxygen species), cell damage from oxidative stress (caused by reactive oxygen or reactive nitrogen species), and local inflammatory effects.25 All of these mechanisms are often interconnected and under various circumstances have been associated with the formation of chemically reactive metabolites.

We also previously evaluated the relationship between daily dosages ≥50 mg against the already assessed FDA hepatic liability categories and DILI severity assessment.8 Our analysis15 concurs with the association of drugs being given at dosages ≥50 mg day–1 having more adverse hepatic events. We have further evaluated this observation by examining the FDA hepatic liability distribution and DILI severity assessment. Drugs with a daily dose ≥50 mg had a much higher frequency of toxicity as evidenced by the higher percentages in the “Warning and Precautions”, “Boxed Warning” and “Withdrawn” label sections (Fig. 2A). For the DILI assessment in Fig. 2B we also observe a higher frequency in DILI severity for compounds that are dosed at ≥50 mg day–1.

Fig. 2. (A) Daily dose ≥50 mg prediction (Safe/Not Safe) vs. FDA Hepatic ADR Categories. (Reprinted with permission from R. Chan and L. Z. Benet, Evaluation of DILI predictive hypotheses in early drug development, Chem. Res. Toxicol., 30, 1017–1029. Copyright 2017 American Chemical Society.) There is a marked increase in the proportion of compounds that are dosed at ≥50 mg day–1 and have FDA drug label warnings associated with DILI adverse effects as illustrated in the “Warnings and Precautions”, “Boxed Warning” and “Withdrawn” categories. (B) Daily dose ≥50 mg prediction (Safe/Not Safe) vs. FDA DILI severity assessment. Similarly, there is a marked increase in the proportion of compounds that are dosed at ≥50 mg day–1 and have some type of DILI toxicity as illustrated in the “Non-severe DILI” and “Severe DILI” categories.

Fig. 2

Open in a new tab

Although, there is strong evidence that dosages ≥50 mg day–1 are associated with increased risk for hepatotoxicity, many drugs are safe at such dosages. For instance, the 50 mg day–1 dosage cut off would predict that 44% of “No mention” and/or “No DILI” drugs (see Fig. 2) exhibit “Not Safe” potential in terms of hepatotoxicity, supporting that daily dosage alone is not a reliable means of guiding the drug development process, regulatory application, and clinical practice.

BDDCS classification prior to dosing in humans

Although the finding of Uetrecht shows that idiosyncratic drug reactions were rare among individuals given drug doses <10 mg day–1 and more likely among individuals given drug doses ≥1000 mg day–1,26 the dose relationships can only be determined for a new molecular entity after the drug has been administered to human subjects/patients. In contrast, BDDCS class can be predicted prior to ever dosing the compounds to animals and humans as we have proposed previously.27 Hosey and Benet28 showed that based on in vitro permeability measurements, the positive predictive value (PPV) for prediction of extensive metabolism were all 90% or greater. And recently Dave and Morris29 showed that they were able to correctly predict highly soluble vs. poorly soluble drugs using measured solubility parameters with greater than 85% probability.

Drug metabolism and propagation of drug hypersensitivity reactions

Drug metabolism also plays an important role in the initiation and propagation of drug hypersensitivity through the generation of neoantigens that are recognized by the cellular and humoral immune systems.2 Although the majority of drug biotransformations occur in the liver, there is overwhelming evidence to suggest that localized drug metabolism by immune cells is critical for organ-specific reactions such as cutaneous adverse drug reactions.2,30,31 These reactions are usually rare and are not typically present in animal species, but they can be serious and even fatal in humans32,33 and may lead to the withdrawal of otherwise effective therapeutic agents. At present, during preclinical drug evaluation there are no widely accepted methods for the identification of drugs that may cause hypersensitivity or idiosyncratic reactions. It has been demonstrated that HLA-B*15:02 is not only a genetic marker but also a key determinant in the pathophysiology of carbamazepine related Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis (SJS/TEN). We have previously assessed the use of the BDDCS to distinguish antiepileptic drugs (AEDs) associated with and without cutaneous adverse events by examining in vitro the binding relationship of AEDs to HLA-B*15:02 via surface plasmon resonance and clinical patient data from extensive reviews of medical records. We also evaluated the lack of benefit from a Hong Kong population policy on the effects of screening for HLA-B*15:02 and previous incorrect structure–activity hypotheses. Our analysis concludes that BDDCS Class 2 AEDs are more prone to cause adverse cutaneous reactions than certain BDDCS Class 1 AEDs and that BDDCS Class 3 drugs have the lowest levels of cutaneous adverse reactions. We proposed that BDDCS Class 3 AEDs should be preferentially used for patients with Asian backgrounds (i.e., Han Chinese, Thai and Malaysian populations) if possible and in patients predisposed to skin rashes.14

Alfirevic and Pirmohamed34 and Urban et al.35 have summarized the current state of pharmacogenomics and suggested that although certain HLA and other differences are related to a higher susceptibility of DILI from a number of agents, the actual number of drugs identified as having these genetic risks is still quite small, and the accuracy of most polymorphisms is limited. Although significant advances in our hepatotoxicity knowledge base have been made by the DILI Network and others,36 when it comes to identifying the specific components of DILI risk, it appears to be much more complex than just being a matter of daily dose or drug disposition.21

Statistical tests of DILI predictivity

Most frequently statistical measures of the performance of binary classification tests are reported as Sensitivity and Specificity. We believe that DILI predictivity deserves a more stipulated analysis. In bringing a new molecular entity (NME) to market, predictive statistical measures for DILI should address (a) how good is the metric in correctly identifying compounds that should be avoided in terms of any further experimental work early in drug development? And (b) what are the chances that an NME showing few preclinical predictors of DILI, will in fact elicit DILI after the company has spent considerable time and money in bringing the drug to market? We believe that the “a” question is best evaluated by the positive predicted value (PPV), the percentage of positive predictions of DILI that actually cause DILI (i.e., True Positive divided by the sum of True Positive and False Positive). If the PPV value for a DILI metric is very high, companies will correctly suspend development of DILI causing compounds very early in the process, and there is only a small chance that a compound not causing DILI will be discarded. We prefer this PPV measure to Specificity (i.e., True Negative divided by the sum of True Negative and False Positive) since we believe PPV more correctly addresses the early issue of stopping development, where Specificity relates to final outcome.

The “b” question is best evaluated by the false negative rate (FNR), the percentage of NMEs that a sponsor takes through extensive and expensive human studies and the regulatory approval process, where DILI manifests after the drug goes on the market. A good DILI predictive metric should show a low FNR (calculated as False Negative divided by False Negative + True Positive). Note the FNR percentage is just 100 minus the Sensitivity percentage. We also list the accuracy (ACC) of the various methods in our tabular comparisons (i.e., Total Positive + Total Negative divided by number of compounds tested). A high PPV metric saves money early in the process by not pursuing compounds highly suspected to cause DILI, but a low FNR metric is probably more important, potentially avoiding very large sunk costs and human suffering after considerable work.

Comparison of in vitro mechanism based toxicity endpoints

Although, a number of compound–specific liability factors have been linked with DILI susceptibility, it is difficult to understand which risk factors are more important in patient-specific responses and/or environmental stimuli. One approach followed by many research groups is to assess and reduce some of the more common, drug-specific factors in a set of targeted in vitro assays. The most common mechanisms covered in in vitro high throughput screening assays include reactive metabolite formation and covalent binding,37,38 inhibition of drug transporters involved in hepatobiliary elimination of bile acids and other metabolic endogenous products (BSEP, MRPs),5,39 mitochondrial toxicity40 and different cellular toxicity assays covering the formation of drug-metabolites.6,41–44 Various approaches are used in the pharmaceutical industry for hazard identification and risk assessment of reactive metabolites and more integrated strategies that include measures of the initial mechanism of toxicity have been highlighted in our analysis.

We have previously performed a comparison of the different predictive metrics in the various assays measuring key mechanisms of toxicity endpoints associated with DILI from the Schadt et al. data set.43 The toxicity endpoints were monitored in a panel consisting of assays assessing the generation of reactive metabolites tested via GSH adduct formation, P450 3A4 time-dependent inhibition (TDI), BSEP inhibition, mitochondrial toxicity and cytotoxicity. In the Schadt et al. data set of 120 marketed compounds, 14 compounds had not been BDDCS classified. Our analysis is depicted in Table 1. (ESI Table S1† includes Sensitivity and Specificity statistical analyses.) The assays that performed the best were GSH adduct formation and BSEP inhibition. We noted that although the positive predictive value (PPV) for these measurements were somewhat better than for BDDCS Class 2 classification, the false negative rate (FNR) for these measures was much greater than BDDCS Class 2, so that in terms of accuracy (ACC), the GSH and BSEP assays were no better than just avoiding BDDCS Class 2 drugs. When GSH and BSEP assays were combined with BDDCS Class 2, again higher PPV values are obtained, but because FNR also increased, ACC is not better than just avoiding BDDCS Class 2 drugs. A slightly higher ACC is obtained when all of the mechanisms of toxicity endpoints are confirmed, due to the low FNR. However, having a PPV of only 65.1% does not give much confidence. We have observed better predictability values for the correlation with MRP3 and MRP4 inhibition, with MRP3 or MRP4 inhibitors giving the best predictability.16 The basolateral efflux transporters, MRP3 and MPR4, play a minor role in bile acid efflux under normal conditions, but they are up-regulated under cholestatic conditions to compensate for impaired biliary excretion. Thus, impaired function of MRP3 and MRP4 by drugs, or genetic polymorphisms resulting in reduced-function variants may result in accumulation of toxic bile acids in hepatocytes.45 But further studies are needed in order to investigate whether this could be a biomarker of value for DILI screening.

Table 1. Comparison of various assays measuring key mechanisms of toxicity endpoints associated with DILI (adapted from ref. 15).

Criteria % Correct (positive predictive value, PPV) % DILI missing (false negative rate, FNR) % Accuracy (ACC) (true positive + true negative)/106
GSH 71.9% 52.1% 69.1%
TDI 75.0% 81.3% 61.8%
Cytotoxicity (3T3 cells) 48.3% 70.8% 55.5%
Mitotox 71.4% 79.2% 61.8%
BSEP 69.2% 62.5% 65.5%
All assays 65.1% 14.6% 73.6%
 
BDDCS Class 1 33.3% 75.0% 45.5%
BDDCS Class 2 64.6% 35.4% 69.1%
 
GSH and BDDCS Class 2 89.5% 64.6% 70.0%
BSEP and BDDCS Class 2 87.5% 70.8% 67.3%
 
BSEP and Mitotox 50.0% 95.8% 56.4%
Open in a new tab

Although there may be some general trends between simple physical parameters, it is unlikely that such considerations could accurately predict risk. This problem could potentially be alleviated by the new in vitro approaches in physiological test systems with model hepatotoxins and utilization of state of the art instrumentation currently being evaluated encompassing chemical and biological factors associated with hepatotoxicity earlier in drug development, as well as the commercial emergence of long-enduring primary hepatocyte-based cultures.15,46,47

Assessment of BDDCS classification on BSEP inhibition and DILI risk

The accumulation of bile acids within hepatocytes is thought to be a primary mechanism for the development of DILI, although as we show with the Schadt et al. data, this is not confirmed. Inhibition of the bile salt export pump (BSEP) by a drug has been implicated as a risk factor for a drug's potential to cause DILI. However, few reports indicate that drug-induced BSEP dysfunction actually leads to hepatotoxicity, and the relationship between drug-induced BSEP dysfunction and liver injury risk is yet to be determined. Recently, the International Transporter Consortium has highlighted BSEP as one of the emerging transporters that need to be considered when evaluating drug safety. However, the practical utility of this approach still needs to be further evaluated. We analyzed further data encompassing the relationship between a compound's ability to inhibit BSEP function in vitro and cause liver injury in humans in vivo using a compilation of published DILI datasets that have screened for BSEP inhibitors, other hepatic transporters and other mechanism based toxicity endpoints such as the mitochondrial toxicity assay.6,43,48,49 We evaluated the information provided by using BDDCS in order to understand the inhibition propensity of BSEP. Our results demonstrate that there is little support for in vitro BSEP inhibition being usefully DILI predictive. Rather we show that the most potent BSEP inhibitors are BDDCS Class 2 drugs, which we have demonstrated previously is the BDDCS class most likely to be DILI related.16

When BSEP inhibition data by Pedersen et al.50 were correlated with the Chen DILI Assessment,51 we observed no discernible pattern16 (see Fig. 3A). For the BDDCS classification, we observe that the great majority of strong BSEP inhibitors are BDDCS Class 2 drugs, with concomitant decreases in the percentages of BDDCS class 1 and 3 drugs as BSEP inhibition increases, as depicted by Fig. 3B. These results along with our analyses of other datasets,16 show that in vitro measures of BSEP inhibition by small molecules is not a strong susceptibility factor. It is unclear as to what extent BSEP inhibition is functionally significant in vivo. Furthermore, our analyses demonstrate that the great majority of compounds that have been associated with DILI and show strong BSEP inhibition potential in vitro are also BDDCS Class 2 drugs. Because we are able to make similar predictions based on BDDCS determinant characteristics, this leads us to discount the predictive ability of in vitro BSEP inhibition for DILI. We have previously observed that as hepatic warning severity increases, the proportion of BDDCS Class 2 drugs increases and the proportions of both BDDCS Class 1 and 3 drugs decrease.15 We conclude that previous analyses predicting that BSEP inhibition leads to DILI may have been confounded by the observation that most BSEP inhibitors are BDDCS Class 2 drugs, which show a high prevalence for DILI.16

Fig. 3. (A) Distribution of BSEP Inhibition with respect to the Chen DILI Assessment (R. Chan and L. Z. Benet, Measures of BSEP inhibition in vitro are not useful predictors of DILI, Toxicol. Sci., 2018, 162, 499–508, by permission of the Royal Society of Chemistry). (120 drugs BSEP Non-Inhibitors, 21 BSEP Weak Inhibitors, and 37 BSEP Inhibitors). (B) Distribution of BSEP Inhibition with respect to BDDCS Class for the Pedersen et al. Data Set49 (Reprinted with permission from R. Chan and L. Z. Benet, Evaluation of DILI predictive hypotheses in early drug development, Chem. Res. Toxicol., 30, 1017–1029. Copyright 2017 American Chemical Society.) (120 drugs BSEP Non-Inhibitors, 22 BSEP Weak Inhibitors, and 39 BSEP Inhibitors).

Fig. 3

Open in a new tab

Aleo et al.6 suggest that mitochondrial toxicity together with BSEP inhibition may provide improved DILI predictability. When we analyzed the predictability of BSEP inhibition together with mitochondrial toxicity, we observe that BDDCS class 2 characterization shows comparable results.16 This is further confirmed in the Schadt et al. data set where we show in Table 1 that combining BSEP inhibition and mitotoxicity yields a very high FNR and no improvement in ACC. Thus, we believe that neither BSEP inhibition nor mitochondrial toxicity are useful independent or combined predictors of DILI.

Studies with long-enduring metabolically competent hepatocyte co-cultures

Many of the deficiencies of the high-throughput DILI screens employed by the pharmaceutical industry based on in vitro measures of hepatocyte toxicity theoretically can be overcome with long-enduring metabolically competent hepatocyte co-cultures. The long-enduring status has the potential to allow formation of active metabolites and/or accumulation of intracellular toxic substances in hepatocytes due to inhibition of relevant efflux transporters, as well as giving sufficient time for the hepatotoxicity to manifest. A December 2017 study47 describes results for 19 identified drugs (12 known to exhibit clinical DILI and 7 that were negative), where measures of TC50/Cmax were obtained in human hepatocyte mono-cultures after 1 day and co-cultures after 1, 7 and 14 days. The results compared to BDDCS classification are summarized in Table 2.

Table 2. Comparison of toxicity endpoints associated with DILI for long-enduring primary based hepatocyte-based co-cultures46 and other measures.

Measurement % Correct (positive predictive value, PPV) % DILI Missing (false negative rate, FNR) % Accuracy (ACC) (true positive + true negative)/19
Hμrel Co-culture Day 1 100% 66.7% 57.9%
Hμrel Co-culture Day 7 100% 33.3% 78.9%
Hμrel Co-culture Day 14 100% 16.7% 89.5%
 
Mono-culture Day 1 100% 50.0% 68.4%
 
BDDCS Class 2 (19 drugs)46 66.7% 20.0% 68.4%
 
Dose ≤50 mg (19 drugs)46 83.3% 16.7% 78.9%
Dose ≤50 mg (historical)7,8 78.3% 9.8% 76.5%
Open in a new tab

None of the seven DILI negative drugs (propranolol, rosiglitazone, diphenhydramine, isoproterenol, kanamycin, macitentan and primidone) gave a positive reading; therefore, all PPV values were 100%. The two clinical DILI drugs (bicalutamide and tacrine) that gave no positive readings under any conditions exhibited decreased TC50/Cmax values with time but did not reach toxic readings. It is interesting to note that BDDCS Class 2 categorization approached the day 1 accuracy values, as seen in the comparisons in Table 2 and our more extensive analyses.15,16 As we have noted previously,14–16 publication of toxicity predictors concentrate in achieving high PPV rates. In contrast, in drug development what is critical to the industry in terms of cost and effort is a predictive metric with a low FNR. In Table 2 we also list the results for our analysis using the criteria of clinical dose ≤50 mg as proposed by Lammert et al.,7 which yielded the lowest FNR of all the data sets we analyzed,15,16 for the 19 drugs investigated by Novik et al.47 and the Chen et al.8 DILI data set for BDDCS classifiable drugs. However, it must be remembered that BDDCS classification and the co-culture analysis reported by Novik et al.47 can occur very early in drug development, certainly before knowledge of the clinical dose and Cmax is obtained. The downside to the Novik et al.47 analysis, based on the ratio of TC50 to Cmax, and the criteria of clinical dose ≤50 mg is that these values will not be known early in drug development.

Yet, even with this limitation, we are encouraged that an additional predictive measure of DILI (in terms of time of incubation) is available. We suspect that if the highly positive results reported by Novik et al.47 for 19 drugs can be confirmed in further studies of many more therapeutic agents, this information will be of great value to drug sponsors and regulators in evaluating Phase 3 results and NDA applications. Furthermore, we expect that positive results from long-enduring hepatocytes in predicting DILI may be combined with advances utilizing in silico predictors of pharmacokinetics and translation of in vitro measures of effective concentrations employing novel biomarker systems facilitating future DILI hypothesis measures that can be shown to be superior to the BDDCS classification as demonstrated in Table 2.

Studies with 3D spherical human liver microtissues (hLiMT) and 2D primary human hepatocytes (PHH)

3D Spherical Human Liver Microtissues (hLiMT) have emerged as promising tools to assess mechanisms of hepatotoxicity, as they demonstrate enhanced liver phenotype, metabolic activity, and stability in culture not attainable with conventional 2D hepatic models. hLiMT are made up of primary hepatocytes and non-parenchymal cells that can be used for repeat-dose long term treatment. Proctor and co-workers investigated the cytotoxicity of 110 marketed drugs in 3D Spherical Human Liver Microtissues (hLiMT) and in 2D primary hepatocytes (PHH) from the same human donor.52 To complement our analysis, we performed a BDDCS classification and evaluated the prediction of these two metabolic-competent hepatocyte cultures on the 100 drugs (of the 110) that had been BDDCS classified. Proctor et al.52 evaluated various thresholds of IC50 among 10, 25, 50 and 100 μM in terms of DILI predictivity as well as margin of safety measures (IC50/Cmax) at the same thresholds for the 3D hLiMT and 2D PHH analyses. In Table 3 we reproduce here the results for the 100 μM threshold for the 100 BDDCS classified compounds using drug-induced cytotoxicity as an endpoint since this threshold exhibited the highest accuracy (the results for all four thresholds together with Sensitivity and Specificity values are given ESI Table S2† for all 110 compounds and in Table S3† for the 100 BDDCS classified compounds). The 3D hLiMT 14 day measures yield lower FNR percentages and higher ACC than the 2D PHH 2 day measures and BDDCS class 2 assignment. However, as seen in Table 3, the best differentiating measurement to predict DILI toxicity is seen for Cmax > 1.3 μM, a parameter that will be influenced by dose size. Although, Proctor and co-workers conclude that the hLiMT demonstrate sufficient capability to warrant exploratory liver injury biomarker investigation, we remain skeptical about the usage of the hLiMT to screen for DILI given that we did not see that any of the proposed criteria had a better prediction than Cmax > 1.3 μM in this large, 100 drug dataset.

Table 3. Comparison of toxicity endpoints associated with DILI for 3D spherical human liver microtissues (hLiMT), 2D primary human hepatocytes (PHH)52 and other measures for the 100 compounds where BDDCS classification was known.

Measurement % Correct (positive predictive value, PPV) % DILI missing (false negative rate, FNR) % Accuracy (ACC) (true positive + true negative)/100
3D hLiMT IC50 cutoff 100 μM 88.6% 37.1% 72.0%
3D hLiMT margin of safety (IC50/Cmax) 100 μM 81.8% 41.9% 66.0%
 
2D PHH IC50 cutoff 100 μM 80.0% 67.7% 53.0%
2D PHH margin of safety (IC50/Cmax) 100 μM 80.6% 59.7% 57.0%
 
BDDCS Class 2 74.3% 58.1% 55.0%
C max > 1.3 μM 83.0% 29.0% 73.0%
Open in a new tab

Why are BDDCS Class 2 drugs more toxic than BDDCS Class 1 drugs?

Several studies have also shown a correlation of total administered dose alone or in combination with drug lipophilicity with higher risk of DILI. However, as we show in Fig. 2 and in our discussion above, dose alone is not able to accurately discriminate all drugs causing DILI. Chen et al.8 proposed a Rule of 2 where PPV for DILI was very high when considering drugs with c log P ≥ 3.0 (calculated lipid water partition coefficient) and dose >100 mg. However, we have shown that the Rule of 2 was slightly less accurate than just BDDCS Class 2 assignment.15 But, using the Lammert et al.7 cutoff of dose >50 mg being more likely to cause DILI, we note that from our compilation of the highest approved dose strength,53 54% of Class 2 drugs exceed 50 mg, while only 33% of Class 1 drugs do so.

Highly lipophilic drugs are cleared by the liver and generally require biotransformation to be eliminated. The parameter c log P may simply be a surrogate for extensive biotransformation and hepatic exposure to a reactive metabolite. If c log P could differentiate DILI potential, we would see equal chances of BDDCS Class 1 and 2 drugs leading to DILI toxicity. As seen in Fig. 1, BDDCS Class 2 compounds predominate among the most severe hepatic toxicities. We also observe a clear differentiation between BDDCS Class 2 and 1 in terms of DILI predictability in Table 1. Furthermore, in our previous analysis,15 we have observed that PPV for c log P ≥ 3 alone is fairly low (76.1% and ACC is 52.3%). Therefore, we do not believe extensive metabolism is an adequate DILI predictor.

A major finding in the development of the BDDCS was the recognition that BDDCS Class 1 drugs, i.e. extensively metabolized, highly permeable, highly soluble, may be shown in vitro to be substrates of both uptake and efflux transporters, but that effects of transporters on BDDCS Class 1 drugs are essentially clinically insignificant in the liver and intestine, as well as the brain. Thus, the unbound concentrations of BDDCS Class 1 drugs in the systemic circulation will reflect unbound concentrations in the liver as well as in the rest of the body, since it is transporters that lead to differences in unbound concentrations in different organs. According to BDDCS53,54 approximately 40% of marketed drugs (i.e., those that are Class 1) will still follow the equivalent free drug concentration hypothesis. However, this will not be true for BDDCS Classes 2, 3 and 4 drugs where transporter effects may lead to different unbound concentrations in the liver and throughout the body. That is, Class 1 drugs will follow the long held assumption in deriving pharmacologic/toxicologic relationships that free drug concentrations are the same throughout the body. But this assumption in pharmacology was made prior to any recognition of the importance of drug transporters in controlling permeability.

It is important to recognize that the compounds evaluated here are drugs that reach the market where sponsors were able to convince the regulatory agencies based on in vitro and preclinical animal studies that toxicity potential, particularly DILI, would be manageable or at least acceptable when the drugs reached the market and were taken by large patient populations as compared to those limited number of patients studied during drug development. Thus, according to our hypothesis, drug company sponsors in their preclinical and clinical studies of Class 1 drugs would be able to reasonably predict drug concentrations in the liver and throughout the body. In contrast, for BDDCS Class 2 drugs, where metabolism is the significant process of elimination, drug concentration measurements in the systemic circulation for these compounds both in the preclinical and clinical studies may poorly predict what concentrations are present in the liver and in other organs of the body. And since it is obvious that DILI occurs more frequently with metabolized drugs, studies in drug development with Class 2 drugs would be poorer predictors of toxicity potential due to the challenges to estimate intracellular concentrations and metabolic processes. Thus, the prevalence of DILI with BDDCS Class 2 drugs could just be circumstantial in that sponsors would be unable to properly evaluate hepatic toxicity for these compounds in designing their clinical studies. This problem could potentially be alleviated by new in vitro approaches and utilization of state of the art instrumentation currently being evaluated.

Conclusions

The application of the BDDCS methodology can help evaluate the potential validity of risk assessment hypotheses. The BDDCS Class 2 susceptibility factor yields similar and in a number of cases better accuracy than the DILI predictive potential biomarkers of other methodologies. Since there is no mechanistic basis for BDDCS Class 2 drugs being most DILI related, if an alternate hypothesis is no more predictive than BDDCS Class assignment, we maintain that the alternate hypothesis is not sufficiently predictive, nor a mechanistic valid hypothesis. As seen in Fig. 1, the BDDCS Class 2 versus Class 1 differentiation only becomes evident with the most severe hepatic toxicities, and then only a 2 : 1 differentiation between BDDCS Class 2 versus Class 1 is found. Lammert et al.'s19 assertion that extensive metabolized compounds are at higher risk of developing DILI can be much improved by differentiating BDDCS Class 2 from BDDCS Class 1 drugs. Daily dosage ≥50 mg alone can only depict a clear relationship with dose with compounds that have been previously associated with DILI, but very limited predictability in differentiating compounds with “No DILI” assignment. There is a general acceptance that BSEP inhibition is a source of toxicity. However, according to our analysis of DILI this is not true for in vitro measures of BSEP inhibition. What we find is that most DILI occurs with BDDCS Class 2 compounds and almost all BSEP substrates and inhibitors are Class 2 compounds, but we observe no relationship between the strength of in vitro BSEP inhibition and toxicity, which makes us believe that the generally held hypothesis is incorrect.

Our review of the BDDCS analysis alongside other DILI toxicity potential biomarkers show that none of the current in vitro methodologies are sufficiently accurate and effective in allowing early identification of new molecular entities that will be DILI free. The comparison of proposed DILI predictive methodology with BDDCS assignment offers a useful tool by which new DILI predictive hypotheses can be evaluated. Thus, we are encouraged by the apparent increased sensitivity, as compared to BDDCS Class 2, for long-enduring metabolically competent hepatocyte co-cultures. Furthermore, using BDDCS classification and finding that a compound is Class 2, one would recognize that priority should be given to more aggressively investigate its DILI potential in mechanistic DILI assays. Some DILI risk factors can be mitigated during the drug design/development process to identify drugs with better chemical attributes with reduced potential to cause human DILI. Hopefully, development of mechanism based toxicity endpoints, such as those previously proposed by Chen et al.51 and Schadt et al.,43 as discussed above, will improve future predictability.

Our review of this work has clearly pointed out that many of the published “predictive DILI” hypotheses do no better than just avoiding BDDCS Class 2 drugs. We propose that comparison of predictive DILI hypotheses with BDDCS class assignment is a useful exercise in determining the relevance of predictive metrics. The results presented herein illustrate how BDDCS can be applied to better understand clinically observed hepatotoxicity and aid in the DILI risk assessment of new molecular entities.

Funding

Rosa Chan was supported by the American Association of Pharmaceutical Scientists Graduate Fellowship, American Foundation for Pharmaceutical Education Pre-Doctoral Fellowship, North American Graduate Fellowship from the American College of Toxicology, Supplemental Training for Education Program from the Society of Toxicology, and NIGMS grant R25 GM56847. Dr Benet is a member of the UCSF Liver Center supported by NIH Grant P30 DK026743.

Conflicts of interest

Dr Benet is a founder, holds stock and is chairman of the scientific advisory board of Hμrel Corporation. Dr Chan declares no conflicts or competing financial interest.

Supplementary Material

Click here for additional data file. (107.6KB, pdf)

Footnotes

†Electronic supplementary information (ESI) available. See DOI: 10.1039/c8tx00016f

References

  1. Shah F., Leung L., Barton H. A., Will Y., Rodrigues A. D., Greene N., Aleo M. D. Toxicol. Sci. 2015;147:500–514. doi: 10.1093/toxsci/kfv152. [DOI] [PubMed] [Google Scholar]
  2. Thompson R. A., Isin E. M., Ogese M. O., Mettetal J. T., Williams D. P. Chem. Res. Toxicol. 2016;29:505–533. doi: 10.1021/acs.chemrestox.5b00410. [DOI] [PubMed] [Google Scholar]
  3. Kaplowitz N. Hepatology. 2013;58:15–17. doi: 10.1002/hep.26295. [DOI] [PubMed] [Google Scholar]
  4. Morgan R. E., Trauner M., van Staden C. J., Lee P. H., Ramachandran B., Eschenberg M., Afshari C. A., Qualls C. W., Lightfoot-Dunn R., Hamadeh H. K. Toxicol. Sci. 2010;118:485–500. doi: 10.1093/toxsci/kfq269. [DOI] [PubMed] [Google Scholar]
  5. Dawson S., Stahl S., Paul N., Barber J., Kenna J. G. Drug Metab. Dispos. 2012;40:130–138. doi: 10.1124/dmd.111.040758. [DOI] [PubMed] [Google Scholar]
  6. Aleo M. D., Luo Y., Swiss R., Bonin P. D., Potter D. M., Will Y. Hepatology. 2014;60:1015–1022. doi: 10.1002/hep.27206. [DOI] [PubMed] [Google Scholar]
  7. Lammert C., Einarsson S., Saha C., Niklasson A., Bjornsson E., Chalasani N. Hepatology. 2008;47:2003–2009. doi: 10.1002/hep.22272. [DOI] [PubMed] [Google Scholar]
  8. Chen M., Borlak J., Tong W. Hepatology. 2013;58:388–396. doi: 10.1002/hep.26208. [DOI] [PubMed] [Google Scholar]
  9. Hoofnagle J. H., Serrano J., Knoben J. E., Navarro V. J. Hepatology. 2013;57:873–874. doi: 10.1002/hep.26175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chen M., Vijay V., Shi Q., Liu Z., Fang H., Tong W. Drug Discovery Today. 2011;16:697–703. doi: 10.1016/j.drudis.2011.05.007. [DOI] [PubMed] [Google Scholar]
  11. Chen M., Bisgin H., Tong L., Hong H., Fang H., Borlak J., Tong W. Biomarkers Med. 2014;8:201–213. doi: 10.2217/bmm.13.146. [DOI] [PubMed] [Google Scholar]
  12. Gleeson M. P. J. Med. Chem. 2008;51:817–834. doi: 10.1021/jm701122q. [DOI] [PubMed] [Google Scholar]
  13. Benet L. Z. Basic Clin. Pharmacol. Toxicol. 2010;106:162–167. doi: 10.1111/j.1742-7843.2009.00498.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Chan R., Wei C., Chen Y., Benet L. Z. AAPS J. 2016;18:757–766. doi: 10.1208/s12248-016-9898-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Chan R., Benet L. Z. Chem. Res. Toxicol. 2017;30:1017–1029. doi: 10.1021/acs.chemrestox.7b00025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Chan R., Benet L. Z. Toxicol. Sci. 2018;162:499–508. doi: 10.1093/toxsci/kfx284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Wu C.-Y., Benet L. Z. Pharm. Res. 2005;22:11–23. doi: 10.1007/s11095-004-9004-4. [DOI] [PubMed] [Google Scholar]
  18. Trontell A. E. Curr. Ther. Res. 2001;62:641–649. [Google Scholar]
  19. Lammert C., Bjornsson E., Niklasson A., Chalasani N. Hepatology. 2010;51:615–620. doi: 10.1002/hep.23317. [DOI] [PubMed] [Google Scholar]
  20. Barton P., Riley R. J. Drug Discovery Today. 2016;21:72–81. doi: 10.1016/j.drudis.2015.09.010. [DOI] [PubMed] [Google Scholar]
  21. Lewis J. H. Clin. Gastroenterol. Hepatol. 2014;12:1556–1561. doi: 10.1016/j.cgh.2014.02.011. [DOI] [PubMed] [Google Scholar]
  22. Chen M., Suzuki A., Borlak J., Andrade R. J., Isabel Lucena M. J. Hepatol. 2015;63:503–514. doi: 10.1016/j.jhep.2015.04.016. [DOI] [PubMed] [Google Scholar]
  23. Srivastava A., Maggs J. L., Antoine D. J., Williams D. P., Smith D. A., Park B. K. Handb. Exp. Pharmacol. 2010;196:165–194. doi: 10.1007/978-3-642-00663-0_7. [DOI] [PubMed] [Google Scholar]
  24. Knowles S. R., Uetrecht J., Shear N. H. Lancet. 2000;356:1587–1591. doi: 10.1016/S0140-6736(00)03137-8. [DOI] [PubMed] [Google Scholar]
  25. Jaeschke H., Gores G. J., Cederbaum A. I., Hinson J. A., Pessayre D., Lemasters J. J. Toxicol. Sci. 2002;65:166–176. doi: 10.1093/toxsci/65.2.166. [DOI] [PubMed] [Google Scholar]
  26. Uetrecht J., Dan L. Annu. Rev. Pharmacol. Toxicol. 2007;47:513–539. doi: 10.1146/annurev.pharmtox.47.120505.105150. [DOI] [PubMed] [Google Scholar]
  27. Broccatelli F., Cruciani G., Benet L. Z., Oprea T. I. Mol. Pharm. 2012;9:570–580. doi: 10.1021/mp2004302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hosey C. M., Benet L. Z. Mol. Pharm. 2015;12:1456–1466. doi: 10.1021/mp500783g. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Dave R. A., Morris M. E. Int. J. Pharm. 2016;511:111–126. doi: 10.1016/j.ijpharm.2016.06.060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Ju C., Uetrecht J. P. J. Pharmacol. Exp. Ther. 1999;288:51–56. [PubMed] [Google Scholar]
  31. Baron J. M., Höller D., Schiffer R., Frankenberg S., Neis M., Merk H. F., Jugert F. K. J. Invest. Dermatol. 2001;116:541–548. doi: 10.1046/j.1523-1747.2001.01298.x. [DOI] [PubMed] [Google Scholar]
  32. Pelkonen O., Raunio H. Environ. Health Perspect. 1997;105:767–774. doi: 10.1289/ehp.97105s4767. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Uetrecht J. P. Drug Metab. Rev. 1992;24:299–366. doi: 10.3109/03602539208996297. [DOI] [PubMed] [Google Scholar]
  34. Alfirevic A., Pirmohamed M. Clin. Pharmacol. Ther. 2012;92:376–380. doi: 10.1038/clpt.2012.107. [DOI] [PubMed] [Google Scholar]
  35. Urban T. J., Shen Y., Stolz A., Chalasani N., Fontana R. J., Rochon J., Ge D., Shianna K. V., Daly A. K., Lucena M. I., Nelson M. R., Molokhia M., Aithal G. P., Floratos A., Pe'er I., Serrano J., Bonkovsky H., Davern T. J., Lee W. M., Navarro V. J., Talwalkar J. A., Goldstein D. B., Watkins P. B. Pharmacogenet. Genomics. 2013;22:784–795. doi: 10.1097/FPC.0b013e3283589a76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Chalasani N., Fontana R. J., Bonkovsky H. L., Watkins P. B., Davern T., Serrano J., Yang H., Rochon J. Gastroenterology. 2008;135:1924–1934. doi: 10.1053/j.gastro.2008.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Nakayama S., Atsumi R., Takakusa H., Kobayashi Y., Kurihara A., Nagai Y., Nakai D., Okazaki O. Drug Metab. Dispos. 2009;37:1970–1977. doi: 10.1124/dmd.109.027797. [DOI] [PubMed] [Google Scholar]
  38. Usui T., Mise M., Hashizume T., Yabuki M., Komuro S. Pharmacology. 2009;37:2383–2392. doi: 10.1124/dmd.109.028860. [DOI] [PubMed] [Google Scholar]
  39. Morgan R. E., van Staden C. J., Chen Y., Kalyanaraman N., Kalanzi J., Dunn R. T., Afshari C. A., Hamadeh H. K. Toxicol. Sci. 2013;136:216–241. doi: 10.1093/toxsci/kft176. [DOI] [PubMed] [Google Scholar]
  40. Porceddu M., Buron N., Roussel C., Labbe G., Fromenty B., Borgne-Sanchez A. Toxicol. Sci. 2012;129:332–345. doi: 10.1093/toxsci/kfs197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Wang Y.-M., Chai S. C., Brewer C. T., Chen T. Expert Opin. Drug Metab. Toxicol. 2014;10:1521–1532. doi: 10.1517/17425255.2014.963555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Xu J. J., Henstock P. V., Dunn M. C., Smith A. R., Chabot J. R., de Graaf D. Toxicol. Sci. 2008;105:97–105. doi: 10.1093/toxsci/kfn109. [DOI] [PubMed] [Google Scholar]
  43. Schadt S., Simon S., Kustermann S., Boess F., McGinnis C., Brink A., Lieven R., Fowler S., Youdim K., Ullah M., Marschmann M., Zihlmann C., Siegrist Y. M., Cascais A. C., Di Lenarda E., Durr E., Schaub N., Ang X., Starke V., Singer T., Alvarez-Sanchez R., Roth A. B., Schuler F., Funk C. Toxicol. In Vitro. 2015;30:429–437. doi: 10.1016/j.tiv.2015.09.019. [DOI] [PubMed] [Google Scholar]
  44. Russmann S., Kullak-Ublick G. A., Grattagliano I. Curr. Med. Chem. 2009;16:3041–3053. doi: 10.2174/092986709788803097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Yang K., Köck K., Sedykh A., Tropsha A., Brouwer K. L. R. J. Pharm. Sci. 2013;102:3037–3057. doi: 10.1002/jps.23584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. LeCluyse E. L., Witek R. P., Andersen M. E., Powers M. J. Crit. Rev. Toxicol. 2012;42:501–548. doi: 10.3109/10408444.2012.682115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Novik E. I., Dwyer J., Morelli J. K., Parekh A., Cho C., Pludwinski E., Shrirao A., Freedman R. M., MacDonald J. S., Jayyosi Z. Toxicol. Appl. Pharmacol. 2017;336:20–30. doi: 10.1016/j.taap.2017.09.013. [DOI] [PubMed] [Google Scholar]
  48. Aleo M. D., Shah F., He K., Bonin P. D., Rodrigues A. D. Chem. Res. Toxicol. 2017;30:1219–1229. doi: 10.1021/acs.chemrestox.7b00048. [DOI] [PubMed] [Google Scholar]
  49. Köck K., Ferslew B. C., Netterberg I., Yang K., Urban T. J., Swaan P. W., Stewart P. W., Brouwer K. L. Drug Metab. Dispos. 2014;42:665–674. doi: 10.1124/dmd.113.054304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Pedersen J. M., Matsson P., Bergström C. A. S., Hoogstraate J., Norén A., LeCluyse E. L., Artursson P. Toxicol. Sci. 2013;136:328–343. doi: 10.1093/toxsci/kft197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Chen M., Borlak J., Tong W. Hepatology. 2016;64:931–940. doi: 10.1002/hep.28678. [DOI] [PubMed] [Google Scholar]
  52. Proctor W. R., Foster A. J., Vogt J., Summers C., Middleton B., Pilling M. A., Shienson D., Kijanska M., Ströbel S., Kelm J. M., Morgan P., Messner S., Williams D. Arch. Toxicol. 2017;91:2849–2863. doi: 10.1007/s00204-017-2002-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Benet L. Z., Broccatelli F., Oprea T. I. AAPS J. 2011;13:519–547. doi: 10.1208/s12248-011-9290-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Benet L. Z. J. Pharm. Sci. 2013;102:34–42. doi: 10.1002/jps.23359. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Click here for additional data file. (107.6KB, pdf)

Articles from Toxicology Research are provided here courtesy of Oxford University Press

RESOURCES