A 2015 focus on preventing drug-induced arrhythmias

ABSTRACT

Drug-induced Torsade de Pointes arrhythmia is a life-threatening adverse effect feared by pharmaceutical companies. For the last decade, the cardiac safety guidelines have imposed human ether-a-go-go-related gene channel blockade and prolongation of QT interval as surrogates for proarrhythmic risk propensity of a new chemical entity. Suffering from a lack of specificity, this assessment strategy led to a great amount of false positive outcomes. Therefore, this review will discuss new pharmaceutical strategies: the cardiac safety proposal that recently emerged, the Comprehensive in vitro Proarrhythmia Assay, combining in vitro assays that integrate effects on main cardiac ion channels, with computational models of human ventricular action potential as well as assays using human stem cell-derived cardiomyocytes for an improved prediction of drug’s proarrhythmic liability, alternative pharmacological perspectives as well as the current treatment of drug-induced long QT syndrome.

KEYWORDS:

• Cardiac arrhythmias
• drugs
• electrophysiology
• cardiac safety
• animal models

Key issues

• Drug-induced long QT syndrome has been closely associated with the blockade of the hERG channel carrying the outward delayed rectifier potassium current (I_Kr). The current preclinical safety guideline ICH S7b imposes the evaluation of repolarization delay related to I_Kr inhibition that are used as surrogates for the assessment of the proarrhythmic risk liability of novel chemical entities.

• Limited especially in the field of specificity, this evaluation strategy needed to be refined. Therefore, a new cardiac safety concept recently emerged for in vitro evaluation of proarrhythmic properties of new chemical entities, the Comprehensive in vitro Proarrhythmic Assay (CiPA): it integrates acute drug effects on main cardiac ion (calcium, sodium and potassium) channels combined with computational modeling (of human ventricular action potential) and human-stem-cell-derived cardiomyocytes, aiming to improve the prediction of proarrhythmic risk liability.

• However, this proposal also has limitations related to the assumptions made from computational modeling of action potential (not capturing beat-to-beat variability of repolarization), drug’s mode of action (determined with acute IC₅₀ on the main cardiac ion currents and channel state-dependent effect not addressed) and human embryonic stem-cell-derived cardiomyocytes that display rather immature electrophysiological phenotype and fetal ultrastructural features. Therefore, further genetic, electrophysiological and pharmacological characterizations along with standardization of experimental protocols will be necessary to limit the variability of response to reference proarrhythmic triggers.

• At present, using automated high-throughput platforms in simplified and immature models, the proposed paradigm seems to be rather suitable for drug lead identification and optimization within a development program than appropriate for a regulatory purpose helping health authorities.

• More importantly, while defective ion channel (hERG) trafficking has been hypothesized to be responsible for most of the drug-induced long QT syndrome, its evaluation upon drug chronic exposure remains, however, uncovered by the proposal.

• Finally, to prevent cardiotoxicity related to the undesired blockade of hERG channel, novel pharmacological strategies modulating hERG channel activity (agonists and allosteric modulators) are currently under development as well as pharmacological rescue of defective hERG channel trafficking.