High-Throughput Screening Assay for Detecting Drug-Induced Changes in Synchronized Neuronal Oscillations and Potential Seizure Risk Based on Ca2+ Fluorescence Measurements in Human Induced Pluripotent Stem Cell (hiPSC)-Derived Neuronal 2D and 3D Cultures.

Drug-induced seizure liability is a significant safety issue and the basis for attrition in drug development. Occurrence in late development results in increased costs, human risk, and delayed market availability of novel therapeutics. Therefore, there is an urgent need for biologically relevant, in vitro high-throughput screening assays (HTS) to predict potential risks for drug-induced seizure early in drug discovery. We investigated drug-induced changes in neural Ca2+ oscillations, using fluorescent dyes as a potential indicator of seizure risk, in hiPSC-derived neurons co-cultured with human primary astrocytes in both 2D and 3D forms. The dynamics of synchronized neuronal calcium oscillations were measured with an FDSS kinetics reader. Drug responses in synchronized Ca2+ oscillations were recorded in both 2D and 3D hiPSC-derived neuron/primary astrocyte co-cultures using positive controls (4-aminopyridine and kainic acid) and negative control (acetaminophen). Subsequently, blinded tests were carried out for 25 drugs with known clinical seizure incidence. Positive predictive value (accuracy) based on significant changes in the peak number of Ca2+ oscillations among 25 reference drugs was 91% in 2D vs. 45% in 3D hiPSC-neuron/primary astrocyte co-cultures. These data suggest that drugs that alter neuronal activity and may have potential risk for seizures can be identified with high accuracy using an HTS approach using the measurements of Ca2+ oscillations in hiPSC-derived neurons co-cultured with primary astrocytes in 2D.

Assessing Drug-Induced Long QT and Proarrhythmic Risk Using Human Stem-Cell-Derived Cardiomyocytes in a Ca2+ Imaging Assay: Evaluation of 28 CiPA Compounds at Three Test Sites.

The goal of this research consortium including Janssen, MSD, Ncardia, FNCR/LBR, and Health and Environmental Sciences Institute (HESI) was to evaluate the utility of an additional in vitro assay technology to detect potential drug-induced long QT and torsade de pointes (TdP) risk by monitoring cytosolic free Ca2+ transients in human stem-cell-derived cardiomyocytes (hSC-CMs). The potential proarrhythmic risks of the 28 comprehensive in vitro proarrhythmia assay (CiPA) drugs linked to low, intermediate, and high clinical TdP risk were evaluated in a blinded manner using Ca2+-sensitive fluorescent dye assay recorded from a kinetic plate reader system (Hamamatsu FDSS/µCell and FDSS7000) in 2D cultures of 2 commercially available hSC-CM lines (Cor.4U and CDI iCell Cardiomyocytes) at 3 different test sites. The Ca2+ transient assay, performed at the 3 sites using the 2 different hSC-CMs lines, correctly detected potential drug-induced QT prolongation among the 28 CiPA drugs and detected cellular arrhythmias-like/early afterdepolarization in 7 of 8 high TdP-risk drugs (87.5%), 6 of 11 intermediate TdP-risk drugs (54.5%), and 0 of 9 low/no TdP-risk drugs (0%). The results were comparable among the 3 sites and from 2 hSC-CM cell lines. The Ca2+ transient assay can serve as a user-friendly and higher throughput alternative to complement the microelectrode array and voltage-sensing optical action potential recording assays used in the HESI-CiPA study for in vitro assessment of drug-induced long QT and TdP risk.

Development of a Human iPSC Cardiomyocyte-Based Scoring System for Cardiac Hazard Identification in Early Drug Safety De-risking.

Human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) have emerged as a promising cardiac safety platform, demonstrated by numerous validation studies using drugs with known cardiac adverse effects in humans. However, the challenge remains to implement hiPSC-CMs into cardiac de-risking of new chemical entities (NCEs) during preclinical drug development. Here, we used the calcium transient screening assay in hiPSC-CMs to develop a hazard score system for cardiac electrical liabilities. Tolerance interval calculations and evaluation of different classes of cardio-active drugs enabled us to develop a weighted scoring matrix. This approach allowed the translation of various pharmacological effects in hiPSC-CMs into a single hazard label (no, low, high, or very high hazard). Evaluation of 587 internal NCEs and good translation to ex vivo and in vivo models for a subset of these NCEs highlight the value of the cardiac hazard scoring in facilitating the selection of compounds during early drug safety screening.

PK/PD Modelling of the QT Interval: a Step Towards Defining the Translational Relationship Between In Vitro, Awake Beagle Dogs, and Humans.

Inhibiting the human ether-a-go-go-related gene (hERG)-encoded potassium ion channel is positively correlated with QT-interval prolongation in vivo, which is considered a risk factor for the occurrence of Torsades de Pointes (TdP). A pharmacokinetic/pharmacodynamic model was developed for four compounds that reached the clinic, to relate drug-induced QT-interval change in awake dogs and humans and to derive a translational scaling factor a 1. Overall, dogs were more sensitive than humans to QT-interval change, an a 1 of 1.5 was found, and a 10% current inhibition in vitro produced a higher percent QT-interval change in dogs as compared to humans. The QT-interval changes in dogs were predictive for humans. In vitro and in vivo information could reliably describe the effects in humans. Robust translational knowledge is likely to reduce the need for expensive thorough QT studies; therefore, expanding this work to more compounds is recommended.

Application of a systems pharmacology model for translational prediction of hERG-mediated QTc prolongation.

Drug-induced QTc interval prolongation (Δ QTc) is a main surrogate for proarrhythmic risk assessment. A higher in vivo than in vitro potency for hERG-mediated QTc prolongation has been suggested. Also, in vivo between-species and patient populations' sensitivity to drug-induced QTc prolongation seems to differ. Here, a systems pharmacology model integrating preclinical in vitro (hERG binding) and in vivo (conscious dog Δ QTc) data of three hERG blockers (dofetilide, sotalol, moxifloxacin) was applied (1) to compare the operational efficacy of the three drugs in vivo and (2) to quantify dog-human differences in sensitivity to drug-induced QTc prolongation (for dofetilide only). Scaling parameters for translational in vivo extrapolation of drug effects were derived based on the assumption of system-specific myocardial ion channel densities and transduction of ion channel block: the operational efficacy (transduction of hERG block) in dogs was drug specific (1-19% hERG block corresponded to ≥10 msec Δ QTc). System-specific maximal achievable Δ QTc was estimated to 28% from baseline in both dog and human, while %hERG block leading to half-maximal effects was 58% lower in human, suggesting a higher contribution of hERG-mediated potassium current to cardiac repolarization. These results suggest that differences in sensitivity to drug-induced QTc prolongation may be well explained by drug- and system-specific differences in operational efficacy (transduction of hERG block), consistent with experimental reports. The proposed scaling approach may thus assist the translational risk assessment of QTc prolongation in different species and patient populations, if mediated by the hERG channel.

Modelling of drug-induced QT-interval prolongation: estimation approaches and translational opportunities.

Safety pharmacology studies are performed to assess whether compounds may provoke severe arrhythmias (e.g. Torsades de Pointes, TdP) and sudden death in man. Although there is strong evidence that drugs inducing TdP in man prolong the QT interval in vivo and block the human ether-a-go-go-related gene (hERG) ion channel in vitro, not all drugs affecting the QT interval or the hERG will induce TdP. Nevertheless, QT-interval prolongation and hERG blockade currently represent the most accepted early risk biomarkers to deselect drugs. An extensive pharmacokinetic/pharmacodynamic (PK/PD) analysis is developed to understand moxifloxacin's-induced effects on the QT interval by comparing the relationship between results of an in vitro patch-clamp model to in vivo models. The frequentist and the fully Bayesian estimation procedures were compared and provided similar performances when the best model selected in NONMEM is subsequently implemented in WinBUGS, which guarantees a straightforward calculation of the probability of QT-interval prolongation greater than 2.5 % (10 ms). The use of the percent threshold to account for the intrinsic differences between species and a new calculation of the probability curve are introduced. The concentration providing the 50 % probability indicates that dogs are more sensitive than humans to QT-interval prolongation. However, based on the drug effect, a clear distinction between species cannot be made. An operational PK/PD model of agonism was used to investigate the relationship between effects on the hERG and QT-interval prolongation in dogs. The proposed analysis contributes to establish a translational relationship that could potentially reduce the need for thorough QT studies.

Sensitivity of pharmacokinetic-pharmacodynamic analysis for detecting small magnitudes of QTc prolongation in preclinical safety testing.

INTRODUCTION: Preclinical concentration-effect (pharmacokinetic-pharmacodynamic, PKPD) modeling has successfully quantified QT effects of several drugs known for significant QT prolongation. This study investigated its sensitivity for detecting small magnitudes of QT-prolongation in a typical preclinical cardiovascular (CV) safety study in the conscious telemetered dog (crossover study in 4-8 animals receiving a vehicle and three dose levels). Results were compared with conventional statistical analysis (analysis of covariance, ANCOVA). METHODS: A PKPD model predicting individual QTc was first developed from vehicle arms of 28 typical CV studies and one positive control study (sotalol). The model quantified between-animal, inter-occasion and within-animal variability and described QTc over 24h as a function of circadian variation and drug concentration. This "true" model was used to repeatedly (n = 500) simulate studies with typical drug-induced QTc prolongation (∆QTc) of 1 to 12 ms at high-dose peak concentrations. Simulated studies were re-analyzed by both PKPD analysis (with varying complexity) and ANCOVA. Sensitivity (power) was calculated as the percentage of studies in which a significant (α = 0.05) drug effect was found. One simulation scenario did not include a concentration-effect relationship and served to investigate false-positive rates. Exposure-effect relationships were derived from both PKPD analysis (linear concentration-effect) and ANCOVA (linear trend test for dose) and compared. RESULTS: PKPD analysis/ANCOVA had a sensitivity of 80% to detect the effects of 7/13 ms (n = 4), 5/10 ms (n = 6) and 4.5/8 ms (n = 8), respectively. The false-positive rate was much higher using ANCOVA (40%) compared to PKPD analysis (1%). Typical drug effects were more precisely predicted using estimated concentration-effect slopes (± 1.5-2.8 ms) than dose-effect slopes (± 3.3-3.7 ms). DISCUSSION: Preclinical PKPD analysis can increase the confidence in the quantification of small QTc effects and potentially allow reducing the number of animals while maintaining the required study sensitivity. This underscores the value of PKPD modeling in preclinical safety testing.

The concordance between nonclinical and phase I clinical cardiovascular assessment from a cross-company data sharing initiative.

It is widely accepted that more needs to be done to bring new, safe, and efficacious drugs to the market. Cardiovascular toxicity detected both in early drug discovery as well as in the clinic, is a major contributor to the high failure rate of new molecules. The growth of translational safety offers a promising approach to improve the probability of success for new molecules. Here we describe a cross-company initiative to determine the concordance between the conscious telemetered dog and phase I outcome for 3 cardiovascular parameters. The data indicate that, in the context of the methods applied in this analysis, the ability to detect compounds that affect the corrected QT interval (QTc) was good within the 10-30x exposure range but the predictive or detective value for heart rate and diastolic blood pressure was poor. These findings may highlight opportunities to refine both the animal and the clinical study designs, as well as refocusing the assessment of value of dog cardiovascular assessments beyond phase 1. This investigation has also highlighted key considerations for cross-company data sharing and presents a unique learning opportunity to improve future translational projects.

Experimentally calibrated population of models predicts and explains intersubject variability in cardiac cellular electrophysiology.

Cellular and ionic causes of variability in the electrophysiological activity of hearts from individuals of the same species are unknown. However, improved understanding of this variability is key to enable prediction of the response of specific hearts to disease and therapies. Limitations of current mathematical modeling and experimental techniques hamper our ability to provide insight into variability. Here, we describe a methodology to unravel the ionic determinants of intersubject variability exhibited in experimental recordings, based on the construction and calibration of populations of models. We illustrate the methodology through its application to rabbit Purkinje preparations, because of their importance in arrhythmias and safety pharmacology assessment. We consider a set of equations describing the biophysical processes underlying rabbit Purkinje electrophysiology, and we construct a population of over 10,000 models by randomly assigning specific parameter values corresponding to ionic current conductances and kinetics. We calibrate the model population by closely comparing simulation output and experimental recordings at three pacing frequencies. We show that 213 of the 10,000 candidate models are fully consistent with the experimental dataset. Ionic properties in the 213 models cover a wide range of values, including differences up to ±100% in several conductances. Partial correlation analysis shows that particular combinations of ionic properties determine the precise shape, amplitude, and rate dependence of specific action potentials. Finally, we demonstrate that the population of models calibrated using data obtained under physiological conditions quantitatively predicts the action potential duration prolongation caused by exposure to four concentrations of the potassium channel blocker dofetilide.

Predicting drug-induced slowing of conduction and pro-arrhythmia: identifying the 'bad' sodium current blockers.

BACKGROUND AND PURPOSE: The regulatory guidelines (ICHS7B) for the identification of only drug-induced long QT and pro-arrhythmias have certain limitations. EXPERIMENTAL APPROACH: Conduction time (CT) was measured in isolated Purkinje fibres, left ventricular perfused wedges and perfused hearts from rabbits, and sodium current was measured in Chinese hamster ovary cells, transfected with Na(v)1.5 channels. KEY RESULTS: A total of 355 compounds were screened for their effects on CT: 32% of these compounds slowed conduction, 65% had no effect and 3% accelerated conduction. Lidocaine and flecainide, which slow conduction, were tested in more detail as reference compounds. In isolated Purkinje fibres, flecainide largely slowed conduction and markedly increased triangulation, while lidocaine slightly slowed conduction and did not produce significant triangulation. Also in isolated left ventricular wedge preparations, flecainide largely slowed conduction in a rate-dependent manner, and elicited ventricular tachycardia (VT). Lidocaine slightly slowed conduction, reduced Tp-Te and did not induce VT. Similarly in isolated hearts, flecainide markedly slowed conduction, increased Tp-Te and elicited VT or ventricular fibrillation (VF). The slowing of conduction and induction of VT/VF with flecainide was much more evident in a condition of ischaemia/reperfusion. Lidocaine abolished ischaemia/reperfusion-induced VT/VF. Flecainide blocked sodium current (I(Na)) preferentially in the activated state (i.e. open channel) with slow binding and dissociation rates in a use-dependent manner, and lidocaine weakly blocked I(Na). CONCLUSION AND IMPLICATIONS: Slowing conduction by blocking I(Na) could be potentially pro-arrhythmic. It is possible to differentiate between compounds with 'good' (lidocaine-like) and 'bad' (flecainide-like) I(Na) blocking activities in these models.

The fentanyl/etomidate-anaesthetised beagle (FEAB) dog: a versatile in vivo model in cardiovascular safety research.

The purpose of conducting cardiovascular safety pharmacology studies is to investigate the pharmacological profiles of new molecular entities (NMEs) and provide data that can be used for optimization of a possible new drug, and help make a selection of NMEs for clinical development. An anaesthetised dog preparation has been used for more than two decades by our department to measure multiple cardiovascular and respiratory parameters and to evaluate different scientific models, leading to more in-depth evaluation of drug-induced cardiovascular effects. An anaesthetic regime developed in house (induction with lofentanil, scopolamine and succinylcholine, and maintenance with fentanyl and etomidate) gives us a preparation free of pain and stress, with minimal effects on the cardiovascular system. This anaesthetic regime had minimal influences on circulating catecholamine levels, on the baroreflex sensitivity, and on all measured basal parameters compared to conscious dogs. All parameters were stable for at least 3 h, with acceptable tolerance intervals, evaluated over 99 safety studies with 3 vehicle treatments (saline, 10% and 20% hydroxypropyl-beta-cyclodextrin). This translates into a highly sensitive model for detecting possible drug-induced effects of NMEs with different mechanisms of action such as: Ca-, Na-, I(Kr)-, I(Ks)-channel blockers, K- and Ca-channel activators, alpha1- and beta-agonists, and muscarinic antagonists. Fentanyl in combination with etomidate is a successful anaesthetic regime in humans [Stockham, R.J., Stanley, T.H., Pace, N.L., King, K., Groen, F. & Gillmor, S.T. (1987). Induction of anaesthesia with fentanyl or fentanyl plus etomidate in high-risk patients. Journal of Cardiothoracic Anesthesia. 1(1), 19-23.]. In the anaesthetised dog, QT correction factors (Van de Water correction and body temperature correction) and risk factors (total, short-term and long-term instability) have been evaluated, using this regime [Van de Water, A., Verheyen, J., Xhonneux, R. & Reneman, R. (1989). An improved method to correct the QT interval of the electrocardiogram for changes in heart rate. Journal of Pharmacological Methods, 22, 207-217.; van der Linde, H.J., Van Deuren, B., Teisman, A., Towart, R. & Gallacher, D.J. (2008). The effect of changes in core body temperature on the QT interval in beagle dogs: A previously ignored phenomenon, with a method for correction. British Journal of Pharmacology, 154, 1474-1481.; van der Linde, H.J., Van de Water, A., Loots, W., Van Deuren, B., Lu, H.R., Van Ammel, K., et al. (2005) A new method to calculate the beat-to-beat instability of QT duration in drug-induced long QT in anaesthetised dogs. Journal of Pharmacological and Toxicological Methods, 52, 168-177.]. Furthermore, this anaesthetic protocol has been used to create different scientific models (long QT, short QT) with different specific end-points (ventricular fibrillation, adrenergic- or pause-dependent TdP) and also their specific precursors: e.g. aftercontractions, phase 2 EADs, phase 3 EADs, DADs, T-wave morphology changes, T-wave alternans, R-on-T, transmural and interventricular dispersion [Gallacher, D.J., Van de Water, A., van der Linde, H.J., Hermans, A.N., Lu, H.R., Towart, R., et al. (2007). In vivo mechanisms precipitating torsade de pointes in canine model of drug-induced long QT1 syndrome. Cardiovascular Research, 76-2, 247-256.]. This paper gives a brief overview of the stability, reproducibility, sensitivity and utility of a well-validated anaesthetised dog model.

Analysis of cross-over designs with serial correlation within periods using semi-parametric mixed models.

The use of semi-parametric mixed models has proven useful in a wide variety of settings. Here, we focus on the application of the methodology in the particular case of a cross-over design with relatively long sequences of repeated measurements within each treatment period and for each subject. Other than an overall measure of the difference between each one of the experimental groups and the control group, specific time point comparisons may also be of interest. To that effect, we propose the use of flexible semi-parametric mixed models, enabling the construction of simulation-based simultaneous confidence bands. The bands take into account both between- and within-subject variabilities, while simultaneously correcting for multiple time point comparisons. Owing to the relatively long sequences of measurements per subject, the presence of serially correlated errors is anticipated and investigated. We illustrate how several formulations of semi-parametric mixed models can be fitted and the construction of simulation-based simultaneous confidence bands using SAS PROC MIXED.

Drug-induced long QT in isolated rabbit Purkinje fibers: importance of action potential duration, triangulation and early afterdepolarizations.

In the present study, we investigated three drug-induced long-QT syndromes in isolated rabbit Purkinje fibers in order to identify the relationship of action potential duration (APD), triangulation of action potentials (APD(90)-APD(40)) and early afterdepolarizations. Isolated rabbit Purkinje fibers were superperfused in Tyrode solution with solvent, indapamide (1 x 10 (-4) M, an I(ks) blocker mimicking long QT1), dofetilide (1 x 10 (-9), 1 x 10 (-8) or 1 x 10 (-7) M, an I(kr) blocker mimicking long QT2) or anthopleurin (1 x 10 (-8) M, an inhibitor of the inactivation of the I(Na(+)) current mimicking long QT3) (n=8 per group) for 25 min, and stimulated at 1 Hz for 20 min and at 0.2 Hz for another 5 min. Indapamide did not change APD and triangulation or elicit early afterdepolarizations even in the presence of beta-adrenergic stimulation with isoproterenol. Dofetilide concentration-dependently prolonged APD(90), increased triangulation and elicited early afterdepolarizations. Anthopleurin markedly increased APD(90) as well as triangulation and elicited early afterdepolarizations. The induction of early afterdepolarizations by dofetilide and anthopleurin was associated with a prolongation of APD(90) or an increase in triangulation, but not with a change in APD(40). Moreover, the degree of the increase in the triangulation was larger than that of APD(90) in long QT2 (dofetilide-induced) and long QT3 (anthopleurin-induced) models in isolated rabbit Purkinje fibers. Our present study indicates that rabbit Purkinje fibers can be used as long QT2 (dofetilide-mimicking) and LQT3 (anthopleurin-mimicking) syndrome models, and confirms that drug-induced long QT1 (indapamide-mimicking) is absent. Our present study also shows the relationship between a prolongation of APD(90) or increase in triangulation and the induction of early afterdepolarizations with dofetilide (I(kr) blocker) and anthopleurin (I(Na) modulator) in isolated rabbit Purkinje fibers.

In vivo measurement of QT prolongation, dispersion and arrhythmogenesis: application to the preclinical cardiovascular safety pharmacology of a new chemical entity.

In addition to in silico and in vitro measurements, cardiac electrophysiology in experimental animals plays a decisive role in the selection of a potential 'cardio-safe' new chemical entity (NCE). The present synopsis critically reviews such in vivo techniques in experimental animals. In anaesthetized guinea-pigs, surface ECG recordings readily identify the typical effects of Class I to IV anti-arrhythmic compounds and of If blockers such as zatebradine on ECG intervals and morphology, but also of non-cardiovascular NCEs affecting cardiac electrical activity via ion channels or neurogenic mechanisms. QT/RR plots indicate that bradycardia is a dominant effect of IKr blockers (dual modulation by IKr of sinus node activity and ventricular repolarization). Nevertheless, correction of QT with Bazett's formula usually distinguishes between drug-induced heart rate reduction and real prolongation of ventricular repolarization (QTc). The anaesthetized guinea-pig model thus is a useful tool for first line in vivo testing of an NCE for effects on cardiac electrophysiology, in particular when combined with measurements of drug levels in plasma and heart tissues. In anaesthetized dogs, advanced ECG analyses identify drug-induced effects on atrial and ventricular intervals, on temporal and transmural dispersion of ventricular repolarization and on incidences of early after-depolarizations. This can be combined with complete haemodynamic, pulmonary and pharmacokinetic analyses in one preparation. However, compound doses/plasma levels needed for effects on ventricular repolarization in this model are substantially higher than those identified in guinea-pigs, at least for IKr blocking compounds. Therefore, we use this 'information-rich' canine model as a second line approach. In awake, trained and appropriately instrumented dogs, readings of surface ECG in combination with cardio-haemodynamic and behavioural assessments can be performed after the administration of an NCE via the expected therapeutic route, including oral medication. However, at higher doses the compound under scrutiny may induce overall behavioural side-effects, related to its primary pharmacological action, such as gastrokinetic repercussions or CNS-mediated sedation or excitation. Such primary pharmacological effects are bound to compromise the evaluation of real drug-induced changes on cardiac electrophysiology, readily identified by resource-friendly setups in smaller animals. Therefore, we use such paradigms as an imperative, final cardiovascular check-up, before a 'First in Man' administration of the NCE. In anaesthetized, methoxamine-challenged rabbits, arrhythmogenic effects of IKr blockers (torsades de pointes) and of dual channel INa/IKr blockers (conduction disturbances) are readily identified. Drug-induced QT dispersion rather than a 'simple' QTc prolongation determines the ventricular arrhythmogenic effect of IKr blockers. The latter effect also depends on the rate of drug delivery (plasma levels vs. heart level, equilibrium throughout the myocardium). Therefore, we use models sensitized for arrhythmogenesis to document further the profile of a comparatively 'cardio-safe' NCE. We conclude that the interpretation of an integrated profile of activity of an NCE on in vitro and in vivo cardiovascular parameters, in comparison with the characteristics of its primary pharmacology and target disease, determines its eventual selection via a scientific, rather than a 'checklist' or 'menu' approach to cardiovascular safety pharmacology. Appropriate tests in experimental animals play a key role in this process.