Investigational New Drug Enabling Nonclinical Safety Assessment of the Iminosugar UV-4, a Broad-Spectrum Host-Targeted Antiviral Agent.

The iminosugar UV-4 is a broad-spectrum antiviral drug candidate with activity in vitro and in vivo against multiple, diverse viruses. The toxicological profile of UV-4, dosed as the hydrochloride salt, was evaluated in single-dose and repeat-dose oral toxicity studies in mice, rats, dogs, and non-human primates (NHP). No moribundity or deaths were associated with the drug up to the maximum tolerated dose. No treatment-related adverse effects were observed following single oral doses in dogs, rats, and mice up to 250, 400, 1000 mg/kg, respectively, and in NHP up to 180 mg/kg administered three times daily for 10 days. UV-4-related findings were generally seen at higher doses after 7- or 14-day exposure. The most common clinical pathology findings (increase in aspartate aminotransferase and decreased platelet count) were consistently found across species and each appeared dose related. The kidney, mesenteric lymph nodes, stomach including gastrointestinal tract, and thymus were identified as target organs in mice, rats, and dogs. In 14-day repeat-dose toxicology studies in mice and dogs conducted in compliance with Good Laboratory Practice regulations, the dog was considered to be the most sensitive species to UV-4 exposure based on the treatment-related adverse effects noted in the identified target organs. The results of these studies demonstrate the safety profile of UV-4 hydrochloride and supported the selection of starting and maximal doses for a single ascending dose first-in-human clinical study.

Pre-clinical study of IRDye800CW-nimotuzumab formulation, stability, pharmacokinetics, and safety.

BACKGROUND: Epidermal growth factor receptor (EGFR) is a target for cancer therapy as it is overexpressed in a wide variety of cancers. Therapeutic antibodies that bind EGFR are being evaluated in clinical trials as imaging agents for positron emission tomography and image-guided surgery. However, some of these antibodies have safety concerns such as infusion reactions, limiting their use in imaging applications. Nimotuzumab is a therapeutic monoclonal antibody that is specific for EGFR and has been used as a therapy in a number of countries. METHODS: Formulation of IRDye800CW-nimotuzumab for a clinical trial application was prepared. The physical, chemical, and pharmaceutical properties were tested to develop the specifications to determine stability of the product. The acute and delayed toxicities were tested and IRDye800CW-nimotuzumab was determined to be non-toxic. Non-compartmental pharmacokinetics analysis was used to determine the half-life of IRDye800CW-nimotuzumab. RESULTS: IRDye800CW-nimotuzumab was determined to be non-toxic from the acute and delayed toxicity study. The half-life of IRDye800CW-nimotuzumab was determined to be 38 ± 1.5 h. A bi-exponential analysis was also used which gave a t1/2 alpha of 1.5 h and t1/2 beta of 40.8 h. CONCLUSIONS: Here, we show preclinical studies demonstrating that nimotuzumab conjugated to IRDye800CW is safe and does not exhibit toxicities commonly associated with EGFR targeting antibodies.

Pre-clinical and cellular toxicity evaluation of 7-methylxanthine: an investigational drug for the treatment of myopia.

The present study entails the toxicity evaluation of 7-methyl xanthine (7-MX), first of its kind molecule found effective in phase II clinical trials for the treatment of myopia, in comparison to other clinically used xanthines i.e., caffeine and theobromine. For acute toxicity evaluation, 7-MX was administered orally in two rodent species (rat and mice) at the doses of 300 mg/kg and 2000 mg/kg and for repeated dose 28-d oral toxicity, at 250, 500, and 1000 mg/kg in rats. Further, cellular toxicity was evaluated in normal breast epithelial (fR2), rat brain C6 glioma (C6 glioma) and human colorectal (Caco-2) cell lines. Also, the cell uptake assay to determine the intestinal permeability of drug was performed in Caco-2 cells. In acute toxicity, 7-MX treatment showed no mortality and toxicity, whereas 66.6% (mice) and 33.3% (rat) mortality was observed in both caffeine and theobromine treatment groups. In repeated dose 28-d oral toxicity, 7-MX treatment was found to have no-observed-adverse-effect level up to the dose of 1000 mg/kg in the present study conducted as per OECD guidelines 407. Also, very high IC50 value of 305.5 and 721 µg/mL was observed for 7-MX in fR2 and C6 glioma cells, respectively. In Caco-2 cells, linear bioavailability and high % cell viability was observed. Thus, 7-MX may be classified as Globally Harmonized System (GHS) category 5 drug with LD50 >2000-5000 mg/kg. Also, the repeated dose 28-d oral toxicity study demonstrated 7-MX to be nontoxic in nature, with cell line toxicity results further endorsing its nontoxic nature.

Comparisons of false negative rates from a trend test alone and from a trend test jointly with a control-high groups pairwise test in the determination of the carcinogenicity of new drugs.

Interest has been expressed in using a joint test procedure that requires that the results of both a trend test and a pairwise comparison test between the control and the high groups be statistically significant simultaneously at the levels of significance recommended in the FDA 2001 draft guidance for industry document for the separate tests in order for the drug effect on the development of an individual tumor type to be considered as statistically significant. Results of our simulation studies show that there is a serious consequence of large inflations of the false negative rate through large decreases of false positive rate in the use of the above joint test procedure in the final interpretation of the carcinogenicity potential of a new drug if the levels of significance recommended for separate tests are used. The inflation can be as high as 204.5% of the false negative rate when the trend test alone is required to test if the effect is statistically significant. To correct the problem, new sets of levels of significance have also been developed for those who want to use the joint test in reviews of carcinogenicity studies.

Preclinical toxicity evaluation of JD5037, a peripherally restricted CB1 receptor inverse agonist, in rats and dogs for treatment of nonalcoholic steatohepatitis.

JD5037 is a novel peripherally restricted CB1 receptor (CB1R) inverse agonist being developed for the treatment of visceral obesity and its metabolic complications, including nonalcoholic fatty liver disease and dyslipidemia. JD5037 was administered by oral gavage at 10, 40, and 150 mg/kg/day dose levels for up to 34 days to Sprague Dawley rats, and at 5, 20, and 75 mg/kg/day dose levels for 28 consecutive days to Beagle dogs. In rats, higher incidences of stereotypic behaviors were observed in 10 mg/kg females and 40 mg/kg males, and slower responses for reflex and sensory tests were observed only in males at 10 and 40 mg/kg during neurobehavioral testing. Sporadic minimal incidences of decreased activity (males) and seizures (both sexes) were observed in rats during daily clinical observations, without any clear dose-relationship. Male dogs at 75 mg/kg during treatment period, but not recovery period, had an increased incidence of gut associated lymphoid tissue hyperplasia and inflammation in the intestine. In both species, highest dose resulted in lower AUCs indicative of non-linear kinetics. Free access to food increased the plasma AUC∞ by ~4.5-fold at 20 mg/kg in dogs, suggesting presence of food may help in systemic absorption of JD5037 in dogs. Based on the study results, 150 mg/kg/day in rats, and 20 and 75 mg/kg/day doses in male and female dogs, respectively, were determined to be the no-observed-adverse-effect-levels (NOAELs).

Predicting the Need for a Tier II Ototoxicity Study From Early Renal Function Data.

History has established that many drugs, such as the antibiotics, chemotherapies, and loop diuretics, are capable of inducing both nephrotoxicity and ototoxicity. The exact mechanisms by which cellular damage occurs remain to be fully elucidated. Monitoring the indices of renal function conducted in the Food and Drug Administration's prescribed set of early investigational new drug (IND)-enabling studies may be the first signs of ototoxicity properties of the new drug candidate. In developing improved and efficacious new molecular entities, it is critically necessary to understand the cellular and molecular mechanisms underlying the potential ototoxic effects as early in the drug development program as possible. Elucidation of these mechanisms will facilitate the development of safe and effective clinical approaches for the prevention and amelioration of drug-induced ototoxicity prior to the first dose in man. Biomarkers for nephrotoxicity in early tier I or tier II nonclinical IND-enabling studies should raise an inquiry as to the need to conduct a full auditory function assay early in the game to clear the pipeline with a safer candidate that has a higher probability of continued therapeutic compliance once approved for distribution.

Imaging mass spectrometry in drug development and toxicology.

During the last decades, imaging mass spectrometry has gained significant relevance in biomedical research. Recent advances in imaging mass spectrometry have paved the way for in situ studies on drug development, metabolism and toxicology. In contrast to whole-body autoradiography that images the localization of radiolabeled compounds, imaging mass spectrometry provides the possibility to simultaneously determine the discrete tissue distribution of the parent compound and its metabolites. In addition, imaging mass spectrometry features high molecular specificity and allows comprehensive, multiplexed detection and localization of hundreds of proteins, peptides and lipids directly in tissues. Toxicologists traditionally screen for adverse findings by histopathological examination. However, studies of the molecular and cellular processes underpinning toxicological and pathologic findings induced by candidate drugs or toxins are important to reach a mechanistic understanding and an effective risk assessment strategy. One of IMS strengths is the ability to directly overlay the molecular information from the mass spectrometric analysis with the tissue section and allow correlative comparisons of molecular and histologic information. Imaging mass spectrometry could therefore be a powerful tool for omics profiling of pharmacological/toxicological effects of drug candidates and toxicants in discrete tissue regions. The aim of the present review is to provide an overview of imaging mass spectrometry, with particular focus on MALDI imaging mass spectrometry, and its use in drug development and toxicology in general.

Nanosuspension formulations of poorly water-soluble compounds for intravenous administration in exploratory toxicity studies: in vitro and in vivo evaluation.

This study was conducted to investigate the use of a nanosuspension for intravenous injection into dogs to increase exposure without toxic additives for preclinical studies in the discovery stage. Nanosuspensions were prepared with a mixer mill and zirconia beads with a vehicle of 2% (w/v) poloxamer 338, which was confirmed to lead to no histamine release in dogs. Sterilized nanosuspensions of poorly water-soluble compounds, cilostazol (Cil), spironolactone (Spi) and probucol (Pro), at 10 mg ml(-1) were obtained by milling for 30 min, followed by autoclaving for 20 min at 121 °C and milling for 30 min (mill-autoclave-mill method). The particle sizes (d50) of Cil, Spi and Pro were 0.554, 0.484 and 0.377 µm, respectively, and the percentages of the nominal concentration were 79.1%, 99.6% and 75.4%, respectively. In chromatographic data, no extra peaks were observed. The particle size of Cil was 0.564 µm after storage for 16 days at 2-8 °C. Cil in nanosuspension, but not in microsuspension, rapidly dissolved in dog plasma. Cil nanosuspension at 0.4 mg kg(-1) and Cil saline solution at 0.03 mg kg(-1) , around the saturation solubility, were intravenously administered to dogs. Nanosuspension increased exposure. The versatility of the mill-autoclave-mill method was checked for 15 compounds, and the particle size of 12 compounds was in the nano range. The nanosuspension optimized in this study may be useful for intravenous toxicological and pharmacological studies in the early stage of drug development. Copyright © 2016 John Wiley & Sons, Ltd.

Ion channel profiling to advance drug discovery and development.

In vitro pharmacological profiling provides crucial information to eliminate drug candidates with potential toxicity early in drug discovery and reduce failure in later stages. It has become a common practice in industry to test lead compounds against a panel of ion channel targets for selectivity and safety liability at early drug discovery stages. Ion channel profiling technologies include binding assays, flux assays, fluorescent membrane potential assays, automated and conventional electrophysiology. Instead of examining compound effects on individual ion channel targets, automated current clamp, optical electrophysiology, and multi-electrode assays have evolved to investigate the integrated compound effects on cardiac myocytes. This review aims to provide an overview of ion channel profiling for cardiac safety and comparisons of various technologies.

Target organ profiles in toxicity studies supporting human dosing: Does severity progress with longer duration of exposure?

We have previously reported the profile of target organs (defined as organs showing histopathological changes) in rodent and non-rodent toxicity studies conducted prior to first-time-in-man (FTiM) for 77 AstraZeneca candidate drugs (CDs). Here, we test the assumption that toxicity is exacerbated by dosing duration by comparing the incidence and severity of target organ toxicities in these ≤ 6 week FTiM studies with those observed in subsequent subchronic/chronic (≥ 3 month) studies. Looking at the effect of dosing duration on severity (pathological score) and incidence (percentage of animals within the group) for the 39 CDs that met the criteria for inclusion (comparable doses between FTiM and subchronic/chronic studies), new toxicities appeared for 31 target organs but existing ones resolved for 29 target organs. Increased severity was more frequent for rodent (16 target organs) than for non-rodent (4 target organs). Most notable changes were a large increase in severity/incidence in liver and in non-rodent lung in contrast to a large decrease in severity and incidence for kidneys/ureter and for the non-rodent thymus. Overall this analysis shows that, even with continued exposure, target organ toxicities of CDs are as likely to show partial or complete recovery as they are to progress in severity.

Target organ profiles in toxicity studies supporting human dosing: an assessment of recovery and chronic dosing.

We have previously reported the profile of toxic effects with respect to target organs (defined as organs showing histopathological changes) observed in rodent and non-rodent toxicity studies conducted prior to first time in man (FTIM) for 77 AstraZeneca candidate drugs (CDs) across a range of therapy areas. The main objectives of the current study were twofold; to determine which target organs observed in the FTIM studies recovered after a dose free recovery period and to determine which additional target organs were observed in subsequent chronic (⩾3month) studies required to support longer term clinical dosing. The analysis showed that ⩾86% of findings in studies supporting FTIM either fully or partially resolved at the end of the recovery period, with profiles of recovery that were similar whether the CD progressed into man or not and across different therapy areas. Compared to observations in FTIM studies, chronic studies identified toxicities in an additional 39% of target organs. Overall these data demonstrate that chronic studies in both rodents and non-rodents provide valuable information for the risk assessment for longer term dosing in humans. In addition, the high levels of recovery demonstrated in this analysis suggest that inclusion of recovery assessments on FTIM studies should be on a case-by-case basis driven by a positive indication of need. This is in line with ICH non-clinical guidance that states that reversibility of severe nonclinical toxicities of potential clinic relevance should be assessed 'when appropriate', but that the evaluation can be based on a study of reversibility or on a scientific assessment.

Standardized UV-vis spectra as the foundation for a threshold-based, integrated photosafety evaluation.

Phototoxicity is a relatively common phenomenon and is an adverse effect of some systemic drugs. The fundamental initial step of photochemical reactivity is absorption of a photon; however, little guidance has been provided thus far regarding how ultraviolet-visible (UV-vis) light absorption spectra may be used to inform testing strategies for investigational drugs. Here we report the results of an inter-laboratory study comparing the data from harmonized UV-vis light absorption spectra obtained in methanol with data from the in vitro 3T3 Neutral Red Uptake Phototoxicity Test. Six pharmaceutical companies submitted data according to predefined quality criteria for 76 compounds covering a wide range of chemical classes showing a diverse but "positive"-enhanced distribution of photo irritation factors (22%: PIF<2, 12%: PIF 2-5, 66%: PIF>5). For compounds being formally positive (PIF value above 5) the lowest reported molar extinction coefficient (MEC) was 1700 L mol⁻¹ cm⁻¹ in methanol. However, the majority of these formally positive compounds showed MEC values being significantly higher (up to almost 40,000 L mol⁻¹ cm⁻¹). In conclusion, an MEC value of 1000 L mol⁻¹ cm⁻¹ may represent a reasonable and pragmatic threshold warranting further experimental photosafety evaluation.

Comparative effects of sodium channel blockers in short term rat whole embryo culture.

This study was undertaken to examine the effect on the rat embryonic heart of two experimental drugs (AZA and AZB) which are known to block the sodium channel Nav1.5, the hERG potassium channel and the l-type calcium channel. The sodium channel blockers bupivacaine, lidocaine, and the l-type calcium channel blocker nifedipine were used as reference substances. The experimental model was the gestational day (GD) 13 rat embryo cultured in vitro. In this model the embryonic heart activity can be directly observed, recorded and analyzed using computer assisted image analysis as it responds to the addition of test drugs. The effect on the heart was studied for a range of concentrations and for a duration up to 3h. The results showed that AZA and AZB caused a concentration-dependent bradycardia of the embryonic heart and at high concentrations heart block. These effects were reversible on washout. In terms of potency to cause bradycardia the compounds were ranked AZB>bupivacaine>AZA>lidocaine>nifedipine. Comparison with results from previous studies with more specific ion channel blockers suggests that the primary effect of AZA and AZB was sodium channel blockage. The study shows that the short-term rat whole embryo culture (WEC) is a suitable system to detect substances hazardous to the embryonic heart.

Orphan drugs, orphan diseases. The first decade of orphan drug legislation in the EU.

PURPOSE: To assess the methodological quality of Orphan Medicinal Product (OMP) dossiers and discuss possible reasons for the small number of products licensed. METHODS: Information about orphan drug designation, approval, refusal or withdrawal was obtained from the website of the European Medicines Agency and from the European Public Assessment Reports. RESULTS: From 2000 up to 2010, 80.9 % of the 845 candidate orphan drug designations received a positive opinion from the European Medicines Agency (EMA)'s Committee on Orphan Medicinal Products. Of the 108 OMP marketing authorizations applied for, 63 were granted. Randomised clinical trials were done for 38 OMPs and placebo was used as comparator for nearly half the licensed drugs. One third of the OMPs were tested in trials involving fewer than 100 patients and more than half in trials with 100-200 cases. The clinical trials lasted less than one year for 42.9 % of the approved OMPs. CONCLUSION: Although there may have been some small improvements over time in the methods for developing OMPs, in our opinion, the number of patients studied, the use of placebo as control, the type of outcome measure and the follow-up have often been inadequate. The present system should be changed to find better ways of fostering the development of effective and sustainable treatments for patients with orphan diseases. Public funds supporting independent clinical research on OMPs could bridge the gap between designation and approval. More stringent criteria to assess OMPs' efficacy and cost/effectiveness would improve the clinical value and the affordability of products allowed onto the market.

Can preclinical drug development help to predict adverse events in clinical trials?

The development of novel therapeutics is associated with high rates of attrition, with unexpected adverse events being a major cause of failure. Serious adverse events have led to organ failure, cancer development and deaths that were not expected outcomes in clinical trials. These life-threatening events were not identified during therapeutic development due to the lack of preclinical safety tests that faithfully represented human physiology. We highlight the successful application of several novel technologies, including high-throughput screening, organs-on-chips, microbiome-containing drug-testing platforms and humanised mouse models, for mechanistic studies and prediction of toxicity. We propose the incorporation of similar preclinical tests into future drug development to reduce the likelihood of hazardous therapeutics entering later-stage clinical trials.

Genotoxicity test battery - An assessment of its utility in early drug development.

Formal requirements for genotoxicity testing of drug candidates to support clinical entry have been in place since the issue of initial regulatory guidance over 25 years ago and subsequent update a decade ago. An evaluation of such testing, supporting first clinical entry of 108 small molecule drug candidates over the last decade, showed that the most common approach (75 % of tested compounds) was for a Good Laboratory Practice test battery in the form of 2 in vitro (a bacterial reverse mutation and a mammalian cell) assays and one in vivo assay. The majority of other tested compounds involved in vitro testing only in bacterial reverse mutation and mammalian cell assays. Testing using a bacterial reverse mutation assay and an in vivo assessment of genotoxicity with 2 different tissues was limited to 2 occasions. For in vitro mammalian cell testing, the chromosome aberration test was most commonly used (70 % occasions), followed by a micronucleus test (16 % occasions) or a mouse lymphoma assay (14 % occasions). For in vivo evaluation, the most common test was a rodent bone marrow micronucleus test (87 % occasions). A positive in vitro mammalian cell assay result was seen on 13 % occasions but was not confirmed with further in vivo testing and the drug candidates were taken into the clinic. In conclusion, the present evaluation showed that the current test battery paradigm for genotoxicity testing has an integral part in supporting clinical entry to confirm candidate drugs taken into the clinic are unlikely to have genotoxic activity.

Current regulatory perspectives on genotoxicity testing for botanical drug product development in the U.S.A.

Genotoxicity testing is an important part of preclinical safety assessment of new drugs and is required prior to Phase I/II clinical trials. It is designed to detect genetic damage such as gene mutations and chromosomal aberration, which may be reflected in tumorigenic or heritable mutation potential of the drug. Botanical new drugs in the U.S. are entitled to a waiver for preclinical pharmacology/toxicology studies, including genotoxicity testing, in support of an initial clinical trial under IND, contingent on previous human experience. Recently, ethical concerns have been raised over conducting Phase I/II clinical trials of new drugs with positive genotoxicity findings in healthy volunteers. Although the relevance of this issue to patients, as opposed to healthy volunteers, depends on the drug's indication, duration of treatment, and specific findings related to the assays, the regulatory view is to avoid exposing patients to genotoxic compounds unnecessarily in clinical trials. This philosophy may impact on herbal supplement marketing and botanical drug development, in that genotoxicity data are often lacking while consumers are exposed to the herbal supplement, or healthy volunteers are tested in an initial Phase I/II clinical trial on the botanical drug. This paper presents results of a survey conducted on genotoxicity data in botanical INDs submitted to the Agency and discusses the significance of this information. The information presented indicates that the sponsors of botanical INDs have increasingly recognized the importance of genotoxicity information and may have prioritized its acquisition in their strategic drug development programs.

Cardiotoxicity Assessment of Drugs Using Human iPS Cell-Derived Cardiomyocytes: Toward Proarrhythmic Risk and Cardio-Oncology.

Growing evidence suggests that Human Induced Pluripotent Stem Cell-Derived Cardiomyocytes (hiPSC-CMs) can be used as a new human cell-based platform to assess cardiac toxicity/safety during drug development. Cardiotoxicity assessment is highly challenging due to species differences and various toxicities, such as electrophysiological and contractile toxicities, which can result in proarrhythmia and heart failure. To explore proarrhythmic risk, the Multi-Electrode Array (MEA) platform is widely used to assess QT-interval prolongation and the proarrhythmic potential of drug candidates using hiPSC-CMs. Several consortiums, including the Comprehensive in vitro Proarrhythmia Assay (CiPA) and the Japanese iPS Cardiac Safety Assessment (JiCSA), have demonstrated the applicability of hiPSC-CMs/MEA for assessing the torsadogenic potential of drug candidates. Additionally, contractility is a key safety issue in drug development, and efforts have been undertaken to measure contractility by a variety of imaging-based methods using iPS-CMs. Therefore, hiPSC-CMs might represent a standard testing tool for evaluating the proarrhythmic and contractile potentials. This review provides new insights into the practical application of hiPSC-CMs in early or late-stage nonclinical testing during drug development.

Enabling speed to clinic for monoclonal antibody programs using a pool of clones for IND-enabling toxicity studies.

The importance of speed to clinic for medicines that may address unmet medical needs puts pressure on product development timelines. Historically, both toxicology and first-in-human clinical materials are generated using the same clonal-derived cells to ensure safety and minimize any development risks. However, cell line development with single cell cloning is time consuming, and aggravated by the time needed to screen for a lead clone based on cell line stability and manufacturability. In order to achieve faster timelines, we have used pools of up to six clones for earlier production of drug substance for regulatory filing-enabling toxicology studies, and then the final single clone was selected for production of clinical materials. This approach was enabled by using platform processes across all stages of early development, including expression vectors, host cell lines, media, and production processes. Through comprehensive cell culture and product quality analysis, we demonstrated that the toxicology material was representative of the clinical material for all six monoclonal antibody programs evaluated. Our extensive development experience further confirmed that using a pool of clones for toxicology material generation is a reliable approach to shorten the early development timeline.

New phenotypic cytotoxicity assay for ROS-inducing compounds using rat renal epithelial cells.

An important mechanism of chemical toxicity is the induction of oxidative stress through the production of excess reactive oxygen species (ROS). In this study, we show that the level of drug-induced ROS production between NRK52E and HepG2 cells is significantly different for several marketed drugs and a number of Takeda's internal proprietary compounds. Nifedipine, a calcium channel blocker and the initial focus of the study, was demonstrated to promote in vitro ROS production and a decrease in cell viability in NRK52E cells but not HepG2 cells. ROS production after nifedipine treatment was inhibited by a NOX inhibitor (GKT136901) but not the mitochondrial NADH dehydrogenase inhibitor, rotenone, suggesting that nifedipine decreases NRK52E cell viability primarily through a NOX-mediated pathway. To understand the breadth of NOX-mediated ROS production, 12 commercially available compounds that are structurally and/or pharmacologically related to nifedipine as well as 172 internal Takeda candidate drugs, were also evaluated against these two cell types. Over 15 % of compounds not cytotoxic to HepG2 cells (below 50 µM) were cytotoxic to NRK52E cells. Our results suggest that a combination of cell viability data from both NRK52E and HepG2 cells was superior for the prediction of in vivo toxicity findings when compared to use of only one cell line. Further, the NRK52E cell viability assay is a good predictor of NOX-mediated ROS production and can be used as a follow up assay following a negative HepG2 response to aid in the selection of suitable compounds for in vivo toxicity studies.