An evaluation of hERG current assay performance: Translating preclinical safety studies to clinical QT prolongation

https://doi.org/10.1016/j.pharmthera.2010.08.008Get rights and content

Abstract

Block of delayed rectifier current (IKr, Kv11.1 encoding the hERG gene) is associated with delayed cardiac repolarization (QTc prolongation), a surrogate marker of proarrhythmia. Despite its recognized role in assessing QTc prolongation risk, a quantitative analysis of the utility and limitations of the hERG current assay has not been reported. To benchmark hERG assay performance, this retrospective study compared hERG block potency with drug-induced QTc prolongation assessed during rigorous thorough QT (TQT) clinical studies for 39 drugs from multiple classes. To place block in context, hERG safety margins (IC50 values for block/mean maximal plasma drug concentrations during TQT studies) were compared to QTc prolongation (QTc increase  5 ms). Most (9/10) drugs eliciting essentially no hERG block at maximal concentrations demonstrate no QTc prolongation despite representing a wide hERG safety margin range. Based on receiver–operator characteristics, a hERG safety margin of 45 provided optimal overall performance linking safety margins to QTc prolongation (sensitivity (true positive rate) = 0.64, specificity (true negative rate) = 0.88); the area under the receiver–operator curve (0.72) is indicative of moderate overall concordance. Likelihood ratios calculated from multitier contingency tables suggest that QTc prolonging drugs are only 5–7 times as likely to demonstrate low safety margins (1–30 range) compared to drugs that do not prolong QTc. Paradoxically, higher safety margins demonstrate lesser confidence predicting prolongation. The overall limitations of hERG safety margins shown using these quantitative, evidence-based approaches highlight the need for additional preclinical assays and adaptive strategies throughout drug discovery to reliably mitigate QTc prolongation risk.

Introduction

Delayed cardiac repolarization is a well-accepted clinical surrogate marker of Torsades-de-Pointes arrhythmia, a rare but potentially lethal ventricular arrhythmia. Numerous studies of acquired long QT syndrome have implicated blockade of the delayed rectifier current IKr, a small but influential cardiac K+ repolarizing current, as a primary factor in promoting pharmacologically-induced delayed repolarization in humans that enhances proarrhythmic risk (see, for example, De Bruin et al. (2005) and Roden (2008)). The human ether-a-go-go related gene (hERG) encodes the pore-forming alpha-subunit of the IKr channel in humans (Sanguinetti et al., 1995). Thus, reduction of IKr/hERG current is considered a surrogate marker of delayed cardiac repolarization, and functional assays measuring drug-induced block of hERG current expressed in heterologous systems are routinely used to characterize the potency of hERG blockade (Gintant et al., 2006, Hancox et al., 2008). Indeed, the value of the hERG assay is recognized by its key role in the integrated risk assessment within the ICH S7b Guidance Document for the Industry (S7b reference) as the primary in vitro assay to consider when evaluating delayed repolarization risk. Despite the generally accepted link between hERG block and delayed repolarization, the relationship between hERG block and clinical QTc prolongation is poorly understood, and reliance on the hERG assay likely results in unwarranted attrition of otherwise promising drug candidates.

In drug discovery, potency of hERG block is typically used to as an early screen for evolving preclinical drug candidates to avoid the risk of delayed cardiac repolarization. The IC50 value for hERG block is often compared to predicted or measured maximal plasma drug concentrations (Cmax) to define a predictive value (IC50/Cmax) referred to as the hERG safety margin. Earlier studies relating hERG safety margins with QTc prolongation and proarrhythmic risk relied on case reports of suspected adverse drug reactions and diverse sets of preclinical and clinical study conditions and impressions (Redfern et al., 2003, De Bruin et al., 2005). While highlighting the utility of the hERG assay and providing provisional safety margins, these studies did not provide a rigorous assessment of the performance of the hERG assay in predicting the extent of small delays in QTc prolongation. Such validation is recognized as an unmet need of safety pharmacology when assessing the risk of untoward clinical effects based on preclinical surrogate markers (Pugsley et al., 2008).

In clinical development, drug-induced delayed cardiac repolarization is rigorously evaluated in “thorough QT” (TQT) studies. A TQT study essentially represents a standardized clinical trial to assess the potential for a compound to affect small changes in ventricular repolarization (manifest as QTc prolongation). A positive control (typically moxifloxacin) is employed to demonstrate study sensitivity (see ICH E-14 document, also Darpo et al. (2006) and Morganroth (2007)). A mean increase in baseline-corrected, time-matched, placebo-controlled, QTc interval of 5 ms is recognized as one indication of concern.

The purpose of this retrospective study was to define the performance of the preclinical hERG current assay in assessing the potential risk of clinical QTc prolongation. This goal was accomplished by comparing TQT study results (prolongation  5 ms, a level of regulatory concern) and hERG safety margins (calculated based on IC50 values for hERG block and mean maximal drug concentrations (Cmax) measured during TQT studies) for 39 drugs. As drug binding to plasma proteins modulates drug distribution, total and free drug concentrations were both considered when defining safety margins. Assay performance was evaluated based on measures of sensitivity, specificity, receiver–operator characteristics, and likelihood ratios (for all approaches for evaluating the utility of diagnostic tests, see Sackett et al. (1991), Jaeschke et al. (1994), Altman and Bland (1994) and Deeks and Altman (2004)). An overview of TQT study results (and moxifloxacin responses from these studies) was also included as part of this review, along with a comparison of hERG block potency to product labeling. Results demonstrate that lower hERG safety margins (values less than 30 based on calculated free drug concentrations) provide moderate confidence in identifying risk of QTc prolongation, with an optimal decision threshold achieved with a safety margin  45. Likelihood ratios (providing another measure of test accuracy) demonstrate that QTc prolonging drugs are 7 times as likely to demonstrate hERG safety margin values below 30 compared to drugs that do not prolong QTc. Paradoxically, higher safety margins (greater than 1000) provide lesser confidence of reduced repolarization risk, an effect likely due to QTc prolongation arising from non hERG-dependent mechanisms and higher drug exposures. While lower safety margins are most useful in identifying the risk of delayed repolarization clinically, neither the potency of hERG block nor hERG safety margins alone are adequate to confidently avoid QTc prolongation. These results provide a strategy for supplementing hERG current assay results with additional preclinical assays early in drug discovery to avoid unwarranted attrition of novel compounds while mitigating delayed repolarization risk.

Section snippets

Methods and background

Information on 39 drugs was obtained from published reports describing hERG and TQT studies, product labels, FDA drug approval packages and EMA approval documents; all information was available in the public domain. No biologics were included in this study, as data regarding potency of hERG current block is often not obtained (as recently supported by a review by Vargas et al. (2008)). For inclusion in this analysis, the following information was required: a) from in vitro studies: IC50 values

Relation of hERG block potency to product labeling

Fig. 1 examines hERG block potency and its relation to product labeling regarding delayed cardiac repolarization. The distribution of IC50 values (based on a 5 μM bin size) for 29 drugs with definitive IC50 values (the standard dataset) is shown in Panel A. The majority of the drugs (72%) were characterized with IC50 values less than 50 μM. Panel B compares block potency for the 39 drug extended datasets with product labeling. Block potency spanned a wide concentration range (~7 log units).

Discussion

This retrospective study of 39 drugs represents the first rigorous quantitative evaluation of the relationship between hERG safety margins (calculated based on hERG block potency in relation to drug exposures during TQT studies) and QTc prolongation assessed during the same TQT studies. As expected, hERG potency alone is inadequate to define either product labeling for QT prolongation or the extent of QT prolongation in thorough QT studies. A general trend relating QTc prolongation with

Acknowledgments

Special thanks to Drs. C. Thomas Lin, Lei Shu and Damian Wandler for statistical discussions, Dr. Anton Safer for early comments, and Dr. Brian F. Hoffman for providing insights still relevant today.

References (43)

  • D.G. Altman et al.

    Diagnostic tests 1: sensitivity and specificity

    BMJ

    (1994)
  • E. Berger et al.

    Investigation of the effects of physiological and vasodilation-induced autonomic activation on the QTc Interval in healthy male subjects

    Br J Clin Pharmacol

    (2005)
  • A. Bril et al.

    Combined potassium and calcium channel blocking activities as a basis for antiarrhythmic efficacy with low proarrhythmic risk: experimental profile of BRL32872

    J Pharmacol Exp Ther

    (1996)
  • L.R. Cantilena et al.

    FDA evaluation of cardiac repolarization data for 19 drugs and drug candidates

    Clin Pharmacol Ther

    (2006)
  • B. Darpo et al.

    Clinical evaluation of QT/QTc prolongation and proarrhythmic potential for nonantiarrhythmic drugs: the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use E14 guideline

    J Clin Pharmacol

    (2006)
  • M.L. De Bruin et al.

    Anti-HERG activity and the risk of drug-induced arrhythmias and sudden death

    Eur Heart J

    (2005)
  • Deeks et al.

    Diagnostic tests 4: likelihood ratios

    BMJ

    (2004, 17)
  • A. Dennis et al.

    hERG channel trafficking: novel targets in drug-induced long QT syndrome

    Biochem Soc Trans

    (2007)
  • L.L. Eckhardt et al.

    Protein trafficking abnormalities: a new mechanism in drug-induced long QT syndrome

    Br J Pharmacol

    (2005)
  • B. Fermini et al.

    The impact of drug-induced QT interval prolongation on drug discovery and development

    Nat Rev Drug Discov

    (2003)
  • C.E. Garnett et al.

    Concentration–QT relationships play a key role in the evaluation of proarrhythmic risk during regulatory review

    J Clin Pharmacol

    (2008)
  • Cited by (171)

    • Improving reliability and reducing costs of cardiotoxicity assessments using laser-induced cell poration on microelectrode arrays

      2021, Toxicology and Applied Pharmacology
      Citation Excerpt :

      Therefore, it is fundamental to improve in vitro methods for assessing accurately the status of all ion-channels in cardiac cells under the influence of the investigated drug. The in vitro measurements of the Ether-à-go-go-Related Gene (hERG) potassium channel activity or QT prolongation have been the traditional standards for preclinical screening assay to evaluate proarrhythmic risk of new drugs(Gintant, 2011) (ICH S7B and E14 guidelines). However, already in 2008, Lu et al.(Lu et al., 2008) questioned the predictivity accuracy of some of the ICH S7B guidelines.

    View all citing articles on Scopus
    View full text